Investing in artificial intelligence is entering a critical testing phase as investors seek tangible evidence of returns following the recent stock market turmoil sparked by disappointing results in the technology sector. This change marks a fundamental shift across the sector from enthusiasm-based to performance-based evaluations.
Oracle shares plummeted 11% Thursday, wiping about $70 billion from its market capitalization after the company reported quarterly revenue that fell short of Wall Street expectations. The database software maker’s revenue of $16.06 billion fell short of the analyst consensus of $16.21 billion despite strong demand for its artificial intelligence infrastructure. This mistake led to a sell-off across AI stocks, including Nvidia, Micron, and Coreweave.
DeVere Group CEO Nigel Green characterized the market reaction as the beginning of a decisive shift that will shape investor behavior from 2026 onwards. After two years of artificial intelligence driving stock performance, the market is drawing a clear line, Green said, and 2026 will be a period of validation as investors look for evidence that big spending programs will lead to sustainable returns.
Oracle’s capital spending totaled about $12 billion in the November quarter, jumping from $8.5 billion in the prior quarter as the company accelerated data center construction to support customers such as OpenAI and MetaPlatform. The outlook for full-year capital spending is $50 billion, up from $35 billion in September, more than double the $21.2 billion in fiscal 2025. Free cash flow turned negative by about $10 billion, compared to analysts’ expectations of a negative $5.2 billion.
The disappointment came even as Wall Street’s composite index hit new record highs. The Dow Jones Industrial Average and Standard & Poor’s 500 both closed at record highs on Thursday, while the Nasdaq Composite Index fell 0.26%. This divergence points to investor selectivity, with money moving away from technology names with uncertain revenue schedules and toward businesses that offer clearer visibility of cash flow.
Artificial intelligence stocks extended their losses into premarket trading Friday, with Oracle down 1.3%, Nvidia and Micron each down 0.9%, and Coreweave down 1.4%. The weakness was the sector’s third straight decline, as skepticism grew over the cyclical nature of some investment agreements and the industry’s large capital requirements.
Oracle’s credit default swaps rose to their highest level since 2009 following the earnings report, according to data from exchange operator ICE. Derivatives serve as a proxy for measuring how risky a company’s bonds and debt appear to market participants. Chief Financial Officer Doug Koehring emphasized on the earnings call that Oracle remains committed to maintaining its investment-grade debt rating despite the surge in infrastructure spending.
Remaining performance obligations, a measure of contract revenue that has not yet been recognized, rose 438% year over year to $523 billion, beating analysts’ expectations of $501.8 billion. The metric reflects a large-scale commitment from large technology companies, but raises questions about execution schedules and funding requirements. Oracle won significant deals during the quarter with customers including Airbus, Canon, Deutsche Bank, London Stock Exchange Group (LSEG), Panasonic and Rubrik.
Green noted that investors are no longer prepared to fund unlimited capital expenditures without a convincing revenue schedule. The questions investors are now asking have changed from how big artificial intelligence can get to how efficiently companies can turn infrastructure, data, and computers into profit. Companies that cannot articulate their path will struggle to justify their valuations in 2026.
This divide is already evident within large technology companies, with some companies showing tighter control over spending and a clearer link between AI adoption and revenue growth. Some companies are facing resistance as shareholders push back against rising costs and uncertain payback periods. As a result, intra-sector diversification has increased, leading to a reassessment of concentration risk and determining selectivity, even as the composite index continues to hit new highs.
By Wednesday’s close, Oracle stock had already lost about 33% of its value since hitting an all-time high on Sept. 10. September’s stock rally followed an earnings call that saw the stock soar 27.5% at one point, based on enthusiasm for the growth of its cloud business and the prediction of large performance obligations. A recent reversal briefly made co-founder Larry Ellison the world’s richest man, but his fortune has fallen along with his stock price.
Policy and geopolitical considerations are placing increased scrutiny on investments in artificial intelligence. As President Donald Trump shapes the current U.S. policy agenda, export controls, domestic supply chain priorities, and national security considerations are influencing how companies plan and invest. These factors are increasingly reflected in valuations and forward guidance as investors consider strategies more broadly.
Analysts point to supply constraints such as graphics processing unit availability, power limitations, liquid cooling challenges and data center lease issues as potential factors that could slow the ambitious build schedule. Oracle sold its stake in chip design company Ampere to SoftBank for $6.5 billion, but Chairman Larry Ellison said the company is no longer considering designing and manufacturing its own chips strategically for cloud data centers.
The dynamics of the broader technology sector reveal increasing interconnectivity through investment agreements. Nvidia announced plans to invest up to $100 billion in OpenAI in September, but Chief Financial Officer Colette Kress acknowledged in December that the final deal was still incomplete. These comments highlight the wide gap between splashy investment announcements and binding contractual obligations across the artificial intelligence ecosystem.
Green concluded that while there is still room for innovation in artificial intelligence, the market is maturing. This realignment comes at a critical time for the sector, after years when enthusiasm alone could support premium valuations. The focus is on operational discipline and profitability as boards and executives face pressure to demonstrate restraint, prioritization, and measurable results.
In industries where artificial intelligence is central to corporate strategy, investment conditions are becoming tougher. Market players expect 2026 to reward companies that demonstrate spending commensurate with revenue potential, credible growth plans, and profitability in line with ambition. The market does not tolerate a disconnect between investment levels and financial performance over time.
For Oracle in particular, the path forward includes optimizing spending on artificial intelligence infrastructure to align with revenue growth. Management expects remaining performance obligations to continue to grow, but Wall Street is looking for evidence of margin improvement and capital efficiency. The company needs to prove that its large data center investments will lead to cloud revenue growth at the pace and scale it promised investors.
Green says the tolerance for assumptions is over and there is a fundamental shift in the way capital markets evaluate artificial intelligence investments. Going forward, companies will need to prove they can maintain financial discipline and execute on ambitious plans while generating returns commensurate with rising valuations. The validation phase, which begins in late 2025, will be extended until 2026, potentially reshaping the competitive landscape within the sector.
