AI Infrastructure Spending Weighs on Big Tech Earnings Outlook, ING Says

Investors are taking a harder look at the soaring costs of artificial intelligence infrastructure, with a new analysis suggesting that massive capital spending could weigh on earnings growth and valuations for major technology companies. The reassessment comes even as underlying demand for AI services remains robust, according to a report from ING.

The Capex Conundrum

The core issue is the significant increase in capital expenditure (capex) required to build out AI capabilities. While necessary to meet demand, this spending is expected to increase depreciation costs, which could slow earnings per share (EPS) growth. This dynamic also reduces free cash flow, potentially limiting the capacity for share buybacks, a key mechanism that has historically supported tech stock valuations.

According to the report, companies like Microsoft (MSFT), Alphabet (GOOGL), and Meta Platforms (META) are particularly noted as likely to see their share repurchase programs constrained by these investments.

Market Volatility and Company Risks

This uncertainty has contributed to significant volatility in tech stocks. In the first half of 2026, for example, Microsoft shares fell 20% and Oracle (ORCL) dropped 27%, while Alphabet gained 14%, reflecting investor ambivalence over the returns on AI spending.

Ad

The analysis highlights several company-specific pressures:

• Microsoft: Is investing approximately $65 billion in cloud and AI infrastructure in fiscal 2025, against an annualized AI revenue of roughly $37 billion. This spending is currently considered economically justified as demand outpaces capacity.
• Oracle: Is identified as a higher-risk case, as its ambitious investment program could strain cash flow and increase its need for external funding.
• Nvidia (NVDA): May face growing long-term competition as key customers including Microsoft, Alphabet, and Amazon (AMZN) develop their own proprietary AI chips to enhance efficiency.

The Justification Question

Despite these pressures, ING views the current level of AI spending as fundamentally sound, driven by the fact that demand for AI computing continues to exceed available supply. Most large tech firms can finance these investments through their substantial operating cash flows.

The central question for investors, however, is whether the eventual revenue growth, profit margins, and earnings expansion from AI will ultimately justify today's level of spending. This uncertainty may lead to investors being less willing to pay premium valuation multiples if AI revenue takes longer than expected to materialize.