The UBS Global Wealth Management Chief Investment Office (CIO) remains bullish on the AI theme, favoring large cap names, leading cloud platforms, and semiconductors globally.
Within China, it believes the internet sector will outperform the broader market amid improving fundamentals, encouraging shareholder returns, and ongoing macro policy support.
“We view AI as the tech theme of the decade, and one that offers significant investment opportunities in the years to come,” said the CIO in a note.
To assess the potential upside that still exists in the AI rally, the CIO uses a framework that considers market capitalisation opportunities with several assumptions.
In global markets excluding China, the CIO forecasts total AI spending to be close to $500bn in 2026, based on current and guided capex commitments from major tech companies.
It expects AI-related revenues, both direct and indirect, to also reach $500bn by then. “As a result, the combined AI end-demand opportunity in 2026 is likely to be close to $1trn,” said the CIO.
Assuming an operating margin of 35%, which is at the lower end of cloud platforms’ margins of 35-40% and AI semiconductors’ 50%, the operating profit opportunity for global AI should stand at around $350bn next year, it added.
To arrive at the market capitalisation opportunity, the CIO then assigns a multiple of 30x next year’s operating profits, in line with levels seen in quality growth companies.
“This would make global AI’s market capitalisation $10.5trn by the end of this year. Looking at what has been priced in so far — $9trn in market capitalisation for global AI names—we see returns of around mid-teens for 2025,” it said.
Nevertheless, the CIO highlights that tariff uncertainty will likely spur periods of tech volatility ahead. But it also sees solid fundamentals from quality AI companies that continue to point to strong investment commitments and improving monetisation trends.
“Investors should consider taking advantage of volatility through structured strategies and by buying the dip in quality AI names,” concluded the CIO.
