Within Western equity markets, two themes are currently stealing the show – artificial intelligence (AI) and defence – with little attention being paid to companies active in other sectors. This concentration on just a few leading names (the “magnificent seven”) is accompanied by excessive valuation multiples. Investors appear to be focusing more on the (evident) disruptive potential of AI than on the (less evident) ability to turn the ongoing AI spending spree into future profits. Which, in some respects, is reminiscent of the late 1990s dotcom bubble.
What exactly is AI ? We apologise if some readers find this a disappointment, but AI is no “new” science – and in fact has little to do with “intelligence”. Rather it is the application of existing mathematical and statistical models to a dataset much larger than the human brain can intuitively process. Indeed, the maximum number of variables that an intelligent, well-educated human brain can handle simultaneously is estimated at seven. It follows that people always make decisions based on incomplete information.
In essence then, AI revolves around processing a far greater amount of data than the human brain can interpret, as well as removing human biases or unconscious choices from decision-making. And the reason for AI’s recent take-off is double: the computing power that IT hardware now offers and the phenomenal amount of money that is being poured into the construction of data centres, i.e. storage facilities for the massive amounts of digital information generated nowadays by internet users/consumers (whether knowingly, or even willingly). Put differently, there is today both the necessary quantity of data and the computing power to enable “AI”.
AI will undoubtedly disrupt the business models of companies across a large range of sectors. Which, from an investor point of view, does clearly make it an important theme. The issue, however, lies in identifying the potential long-term winners. Focussing solely on the current technological leaders may not be the best idea, especially given their current valuations. Moreover, these leaders have built a form of circular ecosystem, whereby they depend on each other to prosper. Chip manufacturer Nvidia is for instance injecting capital into a number of its customer companies (such as in OpenAI for USD 100 billion), in exchange for future chip orders. Similarly, OpenAI recently indicated having made large financial commitments toward suppliers of cloud services and semiconductor manufacturers (USD 1 trillion signed deals in 2025 only).
One lesson from the late 1990s is that many of the most-hyped internet stocks of the time have since disappeared or been relegated an “ordinary” status with standard valuation multiples. So while the internet did live up to its promises, thoroughly shaking up our everyday world for the past quarter of a century, much of the value created was not reaped by the companies that were at the forefront in 2000 and boasted the richest multiples. In hindsight, the long-term internet winners turned out to be, amongst others, the (then non-existent) social media providers.
Another lesson from the late 1990s is that not participating in a hype (bubble) is temporarily very costly in terms of portfolio performance but, conversely, predicting the time at which it will burst is exceedingly difficult, which can lead to very painful consequences for investors.
We worry, thus, about the current valuation multiples reached by stocks such as Nvidia or, worse still, Palantir (because it compounds exposure to AI and defence). Palantir today trades at ca. 300x earnings… And what are we to make of Microsoft’s alleged plans to float OpenAI in a year’s time with a USD 1 trillion valuation in mind, when the company currently generates revenues of only USD 13 billion?
Outside of the AI and defence sectors, the reality is much different. Companies are having to navigate a very uncertain world, with ever-changing (trade) rules, taxes and unsure future consumer demand. Doing business has become a real challenge, and investment plans are tending to be postponed as a result.
For investors, what this boils down to is a limited opportunity set. Within the equity segment, companies with the most interesting earnings prospects have become extremely expensive and the others are, well, rather uninteresting for most investors. In the bond market, European corporate spreads have narrowed sharply, limiting their appeal in our view. There may still be some value to be found among their US peers, but one must then accept the currency risk or hedge it. All told, as we suggested last month already, it may not be a bad idea to focus attention a little more on the eastern part of the globe. And, in any event, portfolios should be kept as diversified as possible.
