Shares of IBM plunged 26% in New York on Tuesday, falling to $215.67 – the largest single-day decline since at least January 3, 1968. The sell-off followed a revenue warning from the company, which blamed soaring costs tied to artificial intelligence infrastructure and severe semiconductor shortages.
IBM reported preliminary second-quarter revenue of $17.2 billion, missing analysts’ estimates of $17.9 billion. The company’s infrastructure division was hit hardest, with sales dropping 7%. Chief Executive Officer Arvind Krishna acknowledged the shortfall in a letter to investors.
– The situation was worse than we expected – said Krishna. – These conditions require flawless execution from our teams, and this quarter we fell short. We failed to adjust fast enough, and many large transactions were not closed in the time we anticipated.
Worse than expected
The market reaction was brutal. IBM’s stock saw its steepest intraday loss in 58 years, and the news dragged down other software firms: Workday and ServiceNow each lost about 6%. In contrast, shares of chip makers such as SK Hynix and Arm Holdings rose, as investors bet on continued demand for AI hardware.
Bloomberg reported that the unprecedented global build-out of data centers – essential for AI systems – has led to severe shortages of semiconductors, especially memory chips. This has pushed up costs for everything from iPads to Xbox consoles. IBM’s results show that companies are being forced to shift spending toward servers and chips, leaving less money for other technologies like mainframes and IBM software.
The real reason behind the plunge
Analysts point out that the problem is not a collapse in AI investment, but a dramatic reallocation of budgets. – In a vacuum, a few percent miss on revenue and profit would justify a significant correction, but not a 26% drop. The real reason lies in the CEO’s comments – said Kamil Szczepański, analyst at XTB. – He openly admits that the company was caught off guard by a sudden change in the structure and timing of client orders.
Szczepański explained that huge capital expenditures fueling the AI revolution are increasingly being diverted to memory chips (DRAM/HBM) due to extreme shortages. This is squeezing companies like IBM that supply software and hardware to the “Big Five” hyperscalers. – It does not mean the AI boom is ending, nor that capex commitments are being cancelled. But it does mean that contracts will be delayed, forcing significant time discounts – he added.
Źródło: wnp.pl, Fot. Tada Images / Shutterstock
