Investors Are Underrating AI, but Diversification Is the Harder Test

Since ChatGPT brought generative AI into everyday conversation, artificial intelligence has dominated financial headlines, earnings calls and boardroom agendas. Yet the global chief investment officer at HSBC believes many investors are still underestimating how deeply the technology will change the economy and the companies that make up the market.

The strategist's concern is less about whether AI matters and more about how portfolios are built around it. Enthusiasm for a theme tends to pile money into a small group of names, which can leave investors far more exposed to one story than they realize. That kind of concentration looks rewarding during a rally, but it offers little cushion if sentiment turns or if the payoff arrives more slowly or unevenly than expected.

Diversification, in this framing, means spreading exposure across the whole chain of activity around AI. That includes the companies building chips and cloud infrastructure, but also the software firms, industrial users and service providers that adopt the technology, and the sectors that could be disrupted by it. A strong theme is not the same thing as a safe bet, and owning the most visible winners does not guarantee that a portfolio is protected.

The message is a balancing act. Investors who ignore AI risk missing a structural shift in how companies grow and compete. Those who chase it without spreading their bets risk the opposite problem. The question for many portfolios is not whether to own exposure to AI, but how much of the portfolio should depend on it, and what happens if that dependence is tested.