AI Spending Payoff? Microsoft Soars 15%, Meta Plunges 9% — Wall Street Delivers Divergent Verdict

Microsoft (MSFT-US) and Meta Platforms (META-US) saw sharply divergent stock movements on Thursday (30th), highlighting investors' contrasting views on large tech companies' artificial intelligence (AI) investments. Microsoft's shares rose 15%, while Meta plunged 9%, extending its record-breaking losing streak.

As of press time, Microsoft's stock was up 15.12%, temporarily trading at $449.80 per share; Meta's stock fell 9.15%, trading at $532.02 per share.

Microsoft reported its fiscal 2026 fourth-quarter earnings on Wednesday, with revenue exceeding analyst expectations. Its key Azure cloud business grew 43% year-on-year in constant currency, surpassing market forecasts. The company also announced that paid seats for Microsoft 365 Copilot have surpassed 30 million, up from over 20 million in April, indicating that some AI investments are beginning to translate into actual revenue.

Tracy Woo, chief analyst at Forrester, noted that Microsoft's strong revenue performance, coupled with accelerating Copilot adoption, suggests that its $190 billion investment in data centers is starting to pay off.

Despite reiterating its capital expenditure outlook for fiscal 2026 and hinting at further spending increases in fiscal 2027, Microsoft's stock surged. Amid growing market concerns over high AI costs, Microsoft's cloud growth and AI product revenue have temporarily eased investor worries about massive capital outlays.

Meta's situation is markedly different. The company's latest quarterly profit missed market expectations, and it forecasted current-quarter revenue between $61 billion and $64 billion. Using the midpoint of $62.5 billion, this falls short of the $63.15 billion average estimate compiled by LSEG.

As Meta continues investing in AI infrastructure, its free cash flow for the quarter plummeted 91% year-on-year to $784 million, deepening market skepticism about return on spending. CEO Mark Zuckerberg revealed that some enterprises are willing to rent computing resources at prices significantly higher than Meta's costs, suggesting the company might lease excess computing power to third parties.

However, Zuckerberg did not elaborate on the business model and admitted Meta must retain sufficient computing power to develop new products. Ben Barringer, Head of Technology Research at Quilter Cheviot, stated that Zuckerberg's current AI vision lacks detail and relies too heavily on future potential outcomes. While Meta will play a key role in AI, it remains in the exploratory phase, and cost and revenue performance are likely to remain volatile.