IBM shares plunged by 20% in premarket trading, and as much as 26% during the day, marking its worst single-day decline since at least 1968, after the company announced preliminary second-quarter revenue of $17.2 billion, significantly below analysts' expectations of $17.86 billion. Adjusted earnings per share are also expected to be lower at $2.93 compared to the forecasted $3.02. This unexpected downturn highlights a critical shift in corporate capital expenditure (capex) priorities.
IBM CEO Arvind Krishna noted that while some supply-chain impacts were anticipated, the "magnitude of the capex reprioritization" was not. He explained that in the last few weeks of June, clients shifted their spending towards servers, storage, and memory purchases to secure supply-constrained infrastructure ahead of expected price increases, diverting funds from software and core IBM offerings, particularly in its mainframe business. Krishna admitted the company "faltered" in adapting quickly enough, leading to numerous large deals not closing as expected, and also cited increased cybersecurity spending due to AI hacking advancements.
The warning from IBM is seen as a significant concern for the broader software industry, as companies that previously invested heavily in software are now reallocating budgets to AI infrastructure. Chris Beauchamp, Chief Market Analyst at IG Group, described it as an "ugly moment for IBM and software stocks," raising questions about the duration of this shift. Other software companies like Microsoft, ServiceNow, Salesforce, and Intuit saw their shares decline between 2% and 5% following IBM's announcement, with the iShares Expanded Tech-Software Sector ETF falling over 4%.
While IBM is making investments in quantum computing ($10 billion to build a large-scale quantum computer by 2029) and expanding AI partnerships, these initiatives are not yet substantial enough to offset the current weakness in its core software and infrastructure businesses. The situation underscores a growing divide where AI infrastructure builders are benefiting, while software companies must adapt to prove how AI will expand, rather than replace, their businesses.
