AI Investor Focus Shifts From Hype to Hyperscaler Fundamentals, Report Says

The investment landscape for artificial intelligence is maturing, with investors shifting from broad enthusiasm to a rigorous analysis of individual company performance. According to a report from Yardeni Research, the market is now scrutinizing the capital expenditure plans, revenue pipelines, and financial health of AI hyperscalers rather than buying into a generalized theme.

Unprecedented Spending Meets Reality

The four largest hyperscalers—Amazon, Alphabet, Microsoft, and Oracle—have laid out combined capital expenditure plans of approximately $600 billion for 2026, a significant increase from $309 billion in fiscal 2025. However, the report notes that delays in power supply, equipment procurement, and local permitting have pushed a substantial portion of this year's planned capacity build-out into 2027 and beyond.

Yardeni Research characterizes this as a situation of "delay, not retreat," pointing to data from Sightline Climate that shows only 9 cancellations out of 777 large-scale data center projects being tracked. Supporting this massive investment is a robust backlog of contracted revenue, known as Remaining Performance Obligations (RPO). The four companies hold a combined RPO of over $2.3 trillion for fiscal 2026, with Microsoft leading at $678 billion and Oracle following at $638 billion.

Funding Pressures and Concentration Risks

This aggressive expansion is introducing new financial pressures. The traditional model of funding capital projects with internal cash flow is changing as spending outpaces free cash flow. Alphabet reported its first-ever quarterly negative free cash flow in the second quarter of 2026, while Amazon's free cash flow for the past 12 months was a net outflow of $7.6 billion.

Ad

Consequently, these firms are increasingly turning to external financing. Citing estimates from Goldman Sachs, the report states that external capital, primarily debt, is projected to fund 33% of hyperscaler CapEx this year, up from 26% in 2025, and is expected to reach about 35% by 2027. The report also highlights customer concentration as a potential risk, noting that S&P downgraded Oracle's credit rating to BBB- in July, citing its significant exposure to OpenAI's ability to fulfill its large compute contracts as a key factor.

A Market of Differentiated Performance

Investors are no longer treating hyperscaler stocks as a monolithic group. The market is now rewarding or penalizing companies based on their specific performance metrics.

• Oracle has seen its stock fall 28% year-to-date amid concerns over its negative free cash flow and heavy reliance on OpenAI.
• Microsoft, after an initial 19% drop, rebounded following strong cloud business results.
• Alphabet and Amazon have performed well, rising approximately 20% and 18% year-to-date, respectively, on the back of strong cloud earnings.

"The AI trade is no longer based on blind faith—as it was in the early days of the AI mania—but on real results," Yardeni Research concluded.