Artificial intelligence capital expenditures could reach $1.6 trillion next year, according to a T. Rowe Price fund manager who compares the current market environment to 1998 rather than the dot-com bust.
The fund manager argues that hyperscalers — the largest cloud and data center operators — can comfortably fund their investment spending without straining their balance sheets. He expects these companies to generate strong returns on invested capital with a relatively short payback period.
The comparison to 1998 is significant because that year preceded the peak of the dot-com bubble. It suggests the AI investment cycle may still have room to grow before any correction occurs.
The manager’s outlook hinges on the ability of hyperscalers to translate massive infrastructure spending into revenue. He believes demand for AI computing power remains robust enough to justify the capital outlay.
Short payback schedules are a key part of the thesis. If AI infrastructure projects deliver returns quickly, companies can reinvest profits into further expansion, creating a self-sustaining cycle.
The $1.6 trillion figure reflects a sharp acceleration from current spending levels. It implies that major technology firms will continue to prioritize AI capacity over other uses of capital.
Skeptics have warned of a potential bubble, drawing parallels to the early 2000s crash. The fund manager, however, sees a more measured trajectory, with fundamentals supporting the spending.
Investors should watch hyperscaler earnings reports closely for signs that AI investments are translating into tangible financial results. Any slowdown in demand could alter the outlook quickly.
For now, the fund manager’s confidence rests on the assumption that AI adoption will keep pace with infrastructure buildout. If that holds, the spending wave could persist well into next year and beyond.
