From the studio

Podcast: What the US-China AI race means for energy markets (9 min)
Video: Our new AI capex forecast and its implications for the AI story (4 min)

Thought of the day

Global stock markets traded edgily on 19 November, with major Asian indexes closing modestly lower, as investors remained cautious ahead of NVIDIA’s earnings results due later today. The company’s partnership with Microsoft—to invest up to a combined USD 15bn in AI developer Anthropic—has also heightened concerns over the circularity of recent AI deals. The S&P 500 fell for a fourth day on 18 November, its longest losing streak in three months.

But despite the recent pullback in tech stocks, we expect AI-driven innovation to propel global stock markets higher. Ongoing investments in AI, the strong financial health of today’s leading tech firms, and both the potential and growing evidence of returns on investments give us confidence in the next leg of the global equity rally in the months ahead.

Robust AI capex may continue to surpass expectations. Strong capital expenditure (capex) has been the biggest driver of AI performance so far, and we believe it will remain robust for longer. Megacap tech firms have recently further raised their spending plans in the coming year, and we now expect global AI capex to reach USD 571bn in 2026, a 35% increase from our estimate of USD 423bn for 2025. By 2030, we expect annual spending to hit USD 1.3tr, implying a compound annual growth rate (CAGR) of 25% between now and then. Based on how capex estimates have been exceeded over the past few years, and how demand for compute could accelerate further as AI tasks become more complex, our projections could ultimately prove conservative.

Big tech companies have sufficient funds to finance their AI expansion. Despite growing capex, most big tech companies today generate substantial operating cash flows that more than cover these investments. The recent wave of bond issuance by these companies, in our view, is more of a financing choice than a need, and we view the current mix of funding approaches as still healthy. While NVIDIA’s recent partnerships have been compared with vendor financing practices prevalent during the dotcom era, today’s deals are subject to stringent disclosure requirements and enhanced accounting standards. The scale of vendor financing has also declined significantly. Following the latest deal with Anthropic, we estimate NVIDIA’s recent collaborations account for only 10% of its projected pretax earnings for 2026, well below the over 120% levels observed during the late 1990s.

Growing evidence of AI monetization bodes well for future growth. While AI monetization has so far lagged capex, we expect the gap to narrow as big tech companies continue to scale monetization. Leading cloud platforms have reported accelerating revenue growth, and management teams have highlighted growing monetization potential with innovative AI tools. AI is also driving productivity gains, with companies reporting tangible value creation and daily time savings for employees. Overall, we continue to believe that the monetization potential for AI is large, even compared to the substantial capex plans.

So, without taking any single-company view, we believe sufficient exposure to AI-linked stocks is critical in building and preserving wealth over time.