OPINION: Are we in an AI bubble?

This exceptional performance has sparked frequent client discussions around the sustainability of the artificial intelligence (AI)-driven rally. Is the surge in US technology stocks fundamentally justified, or does it increasingly resemble a classic bubble? While some areas of the AI space may exhibit signs of froth, suggesting potential consolidation ahead, we see important distinctions between today’s tech rally and previous market bubbles.

We continue to maintain exposure to these companies within our portfolios, but we remain focused on ensuring that overall returns are not overly reliant on the fortunes of this single segment of the market.

Unlike the early 2000s, when valuations soared largely on euphoria, today’s leading technology companies have delivered exceptional earnings growth over recent years.

A simple, but representative, example is the contrast between Cisco during the dot-com era and Nvidia today. Cisco’s share price rise far outpaced its fundamentals, whereas Nvidia’s remarkable price increase has been accompanied by equally impressive profit growth.

Today’s tech giants stand on solid financial ground, supported by diversified revenue streams, mature business models, and fundamentals-driven valuations. For example, their combined net income margins average around 30%, nearly twice that of dot-com firms in the early 2000s.

Equally important, their capital expenditure is largely financed through free cash flow rather than debt, marking a stark contrast to the dot-com era, when many companies lacked profitability and relied heavily on debt or equity issuance to fund growth, leaving them highly exposed when market conditions deteriorated.

Valuations are high, but not extreme relative to history

The Magnificent Seven trade at a median 24-month forward price-to-earnings (P/E) ratio of 27 times expected earnings, or 26 times if Tesla is excluded. P/E is a valuation metric that compares a company’s current share price to its expected earnings per share (EPS) over the next 12 months. Higher P/Es indicate investors expect strong growth or the stock may be expensive relative to earnings.

This is about half the late-1990s tech peak and a third of Japan’s bubble-era highs. Yes, valuations are rich, but so are most major equity and credit markets, driven by low rates and loose monetary policy experienced post Global Financial Crisis.

Ultimately, valuations matter for long-term returns; investing when valuations are low tends to deliver stronger long-term returns than buying into expensive markets. However, valuation alone is rarely a reliable timing tool.

Is there anything to worry about?

Although we don’t believe we’re necessarily in a bubble, risks on the horizon always remain. Key factors that could disrupt the current tech bull market super cycle include:

· Capital investment outpacing return on investment? US tech giants are set to pour nearly $400 billion into capital expenditures this year. Investors have cheered, but the payoff is far from certain. History offers a cautionary tale: in the late 1990s for instance, telecom firms spent hundreds of billions on fiber-optic networks, betting on explosive internet growth. Their optimism proved costly, leaving the industry mired in excess capacity and collapsing prices for years.

· Rising competition? Historically, the technology sector has seen new entrants challenge incumbents and eventually dominate emerging products and technologies, especially in the US For perspective, only about 10% of Fortune 500 companies have remained on the list since 1955, and just one company from the 2000 list was still in the top 10 by 2020. This offers a clear reminder of why diversification matters.

So, what should investors do?

US mega-cap tech stocks look far more reasonably valued today than during the dot-com bubble, but complacency is dangerous. AI demand may be structural and long-lasting, but investor appetite for risk remains cyclical. That’s why diversification matters: markets move in cycles, and what wins today may falter tomorrow.

Take history as a guide: In the 1990s: The S&P 500 (in GBP terms) soared ~18% annually. This is like drafting a star player who dominates every game. In the 2000s: The “lost decade” followed, with negative returns. This is like that same star getting injured in Week 1 and missing the season. Our mission? Avoid periods of feast and famine.

Within our multi-asset products we aim to build ‘sleep-at-night’ portfolios, globally diversified, resilient across market conditions, and never reliant on a single trend. Yes, we maintain exposure to leading tech names, but we ensure portfolios aren’t hostage to one segment’s fortunes.