Asset manager: Seven AI stocks are driving the market, but a bubble hasn't been ruled out

AI is taking up more and more space in the financial markets. That is the view of asset manager Hamilton Lane, which in a 2026 market review describes how the technology is shaping both public exchanges and private investments. Mario Giannini, executive co-chairman of the firm, calls AI a decisive market force.

A few stocks are carrying the load

According to Hamilton Lane, seven AI-related stocks have driven returns in the public markets, meaning the ordinary stock exchanges, for more than six years. The firm believes that will likely hold for the next 2-3 years as well.

Giannini points out that the public equity markets have become concentrated around hyperscalers (the very largest providers of cloud and computing power) and companies with exposure to large language models, like those behind chatbots.

Hamilton Lane also reports that more than half of the value in venture capital deals now goes to AI-focused companies. Venture capital is money invested in young companies with growth potential.

Private markets as an alternative

Giannini emphasizes that private equity provides access to a large market of unlisted companies. These markets do not have the same concentration in big tech companies as the public markets. He stresses that AI doesn't only affect tech stocks, but also credit markets, infrastructure, real estate and energy.

Is it a bubble?

Hamilton Lane puts the probability that there is no AI bubble at about 60%. The firm bases this on the fact that the investments are being made by large, financially strong players, and that physical constraints are dampening overinvestment. According to the firm, it is currently construction time for data centers and energy supply, not demand, that sets the limit.

Giannini is more cautious on one point, however. He warns of uncertainty about the ultimate returns on AI infrastructure and the risk of low capacity utilization if AI usage doesn't live up to expectations. He singles out energy and data centers in unconventional markets.

Hamilton Lane recommends building a portfolio that can withstand both outcomes, bubble and no bubble. The firm points out that stagnation in language models can easily coexist with thriving AI applications.

Hamilton Lane itself believes no bubble is the most likely outcome, but Giannini also points to uncertainty about returns on AI infrastructure. Both sources are Hamilton Lane's own or closely affiliated communications and therefore reflect the firm's views and business interests.