From Oracle to Broadcom, AI Stocks Face Growing Skepticism and Reevaluation
Artificial intelligence has been a major driver of stock-market performance over the past few years, but concerns about AI stocks and their sky-high valuations aren’t new. After a long period of gains, investors are no longer willing to overlook questions about sustainability. From Oracle to Broadcom, the air is filling with more caution as buyers scrutinize whether the enthusiasm can last.
The hurdle now appears to be the practicality of those lofty expectations
Rising skepticism around AI stocks has been building: stretched valuations, hefty capital expenditures, and doubts about whether the world has the capacity to meet surging AI demand. Data centers, the energy they consume, and even essential resources like water for liquid cooling are all part of the feasibility puzzle.
Oracle’s September earnings showcased the upside: the company reported over $450 billion in remaining performance obligations, driven by a fast-growing AI cloud-services business that rents data-center capacity with GPU clusters to power AI workloads. The enthusiasm was evident as the stock jumped about 40% on the strength of demand from hyperscalers such as Microsoft and OpenAI. Yet, momentum cooled in the following months as headlines suggested Oracle would need around $38 billion in new debt to fund data-center expansion, and as some analysts noted that the AI-data-center business, while growing, was delivering margins in the 10%–20% range—lower than some investors had hoped.
In Oracle’s latest quarterly performance for the second fiscal quarter of 2026, the period ending November 30, revenues neared $16.1 billion but barely met consensus estimates, and the stock pulled back. The company also raised its full-year capital expenditure outlook to about $50 billion from $35 billion and reported negative free cash flow of around $10 billion for the quarter. The discomfort extended into credit markets, where five-year credit-default swaps on Oracle’s debt touched new records, reflecting investor concern about risk and future default protection costs.
Broadcom’s results added to the tension. Although Broadcom delivered stronger-than-expected actual results for its fourth-quarter 2025, ending November 2, management’s guidance suggested margins would contract next quarter, and the market felt the AI product backlog might be understated. Even with solid near-term numbers, the longer-term outlook raised questions about how quickly AI-related revenue would scale and how much of the backlog would convert to realized profits.
The overarching issue is whether massive ongoing and planned AI-capital expenditures by hundreds of billions of dollars at large corporations—and the trillions of dollars projected to flow into AI over the coming years—will actually yield meaningful returns.
IBM CEO Arvind Krishna has argued that the math doesn’t support a universal view of outsized returns on data-center investments. He noted that building a 1-gigawatt data center today costs about $80 billion; extrapolate that to 20–30 gigawatts for a single company, and you’re looking at roughly $1.5 trillion in capex. Such estimates highlight the enormity of the challenge and raise serious questions about payback timelines.
Perhaps it’s time to pause and reassess. There’s broad agreement that AI can transform many aspects of industry and society, but the path forward is increasingly uncertain. Resource constraints and debt levels may temper expectations, especially for returns on timelines investors have in mind.
As with the early days of the internet boom, a period of sober evaluation often follows exuberant growth. This could be an opportune moment for investors to reexamine AI holdings, scrutinize valuations, and interrogate the assumptions underpinning growth forecasts. If something looks too good to be true, it may be prudent to scale back gains and rebalance risk, rather than chase further headlines.
Would you adjust your AI exposure in light of these concerns, or do you believe the long-term opportunity justifies current prices? Share your thoughts in the comments.
