IBM shares cratered more than 25% on Tuesday. The drop erased roughly $70 billion in market value. It stands as the company’s worst single-day performance since 1968.
Investors didn’t see it coming. Neither, it seems, did IBM’s own leadership. The tech giant released preliminary second-quarter results a full week early. Revenue came in at $17.2 billion. That marked just 1% growth from the prior year. Adjusted earnings per share hit $2.93. Both figures missed Wall Street forecasts.
CEO Arvind Krishna pointed to a sharp shift in customer spending. Enterprises poured money into AI servers, storage and memory chips. They pulled back from traditional software and infrastructure projects. IBM’s infrastructure segment fell 7%. Software revenue rose 5%. But the overall mix disappointed.
“This quarter we faltered,” Krishna wrote in a letter to investors, according to a Yahoo Finance report. “We did not adapt and move quickly enough.” He added that while the company expected some supply-chain pressure, “we did not anticipate the magnitude of the capex reprioritization.” And these realities, he noted, are “not excuses.”
But the market delivered its verdict anyway. The stock’s slide outpaced even the 23.7% tumble IBM suffered on Black Monday in 1987. Shares traded near $220 in the aftermath. That put the company down 26% for the year to date.
The pain spread beyond IBM. Oracle, Microsoft and Accenture have faced pressure this year too. Memory shortages ripple across the industry. Apple, Sony and Nintendo feel the pinch. Demand for high-bandwidth memory chips outstrips supply as data centers race to power generative AI models.
IBM had banked on its z17 mainframe to drive growth. The system handles massive transaction volumes. It layers in AI for real-time fraud detection. Credit card swipes, ATM withdrawals and stock trades run on it daily. The launch was supposed to be the strongest in the product’s history. Instead, buyers delayed deals. Several large contracts slipped past quarter’s end.
Watson’s Long Road to Relevance
IBM’s watsonx platform was meant to anchor the company’s AI push. It promised enterprise-grade tools for data governance, model training and deployment in hybrid cloud settings. Yet the preliminary results barely mentioned it. Analysts wonder if clients view watsonx as table stakes rather than a must-have differentiator.
Recent coverage highlights the gap. A TIKR analysis from June 2026 flagged investor concerns after IBM’s own AI sovereignty study. The research, conducted with Oxford Economics, found 91% of executives lack full visibility into their AI dependencies. Another 71% said switching primary AI vendors would prove difficult. That combination can breed hesitation, not acceleration.
Customers appear to favor pure-play infrastructure providers for now. NVIDIA dominates the GPU conversation. Hyperscalers build their own silicon. IBM’s hybrid-cloud message resonates with boards worried about lock-in. But when budgets tighten, hardware that directly trains large language models wins the day.
So what happens next? IBM schedules its full second-quarter report for July 22. Management had guided for software growth above 10% for the full year. Holding that line would signal watsonx and recent acquisitions like Confluent still have momentum. A slip toward mid-single digits would confirm the spending rotation runs deeper than expected.
History offers mixed lessons. IBM reinvented itself before. It exited personal computers. It doubled down on services. It acquired Red Hat to bolster cloud credentials. Each pivot carried risk. This time the stakes center on artificial intelligence. The company saved $4.5 billion internally through its own AI initiatives over three years, a Yahoo Finance video segment in June 2026 reported. Jonathan Adashek, IBM’s senior vice president of marketing and communications, described how automation freed creative teams and sharpened sales leads.
That client-zero success story hasn’t translated to the broader market yet. Enterprise buyers cite integration complexity. They question whether watsonx matches the velocity of open-source alternatives or the raw power of cloud-native AI services.
Wall Street reactions poured in fast. CNBC detailed the earnings warning and its immediate fallout. Axios called the move “rattling” for many investors. Dow Jones data noted the plunge put IBM on track for a record percentage decline. Each account stressed the same theme. The AI spending boom disrupts legacy tech budgets faster than incumbents predicted.
But disruptions create openings too. IBM’s installed base remains massive. Banks, insurers and governments run mission-critical workloads on its hardware and software. If the company can demonstrate watsonx’s ability to govern AI across those environments, it might recapture momentum. Governance, compliance and explainability matter more than ever as regulators scrutinize generative tools.
For now the numbers tell a sobering tale. One bad quarter doesn’t redefine a 114-year-old company. Yet it exposes vulnerabilities in timing and forecasting. Krishna’s letter conceded as much. Adaptation must accelerate. The memory shortage won’t last forever. When supply catches up, will customer dollars flow back toward IBM’s full stack? Or has the AI arms race permanently redrawn the competitive map?
Investors will parse every word on July 22. They’ll watch software bookings. They’ll scrutinize mainframe traction. And they’ll measure how credibly IBM positions watsonx against an array of hungry rivals. The stock may have already priced in the worst. Recovery, if it comes, will hinge on execution. Nothing less.
WebProNews is an iEntry Publication