IBM crashes 25% after unscheduled earnings warning, erasing $68B in market cap

Author: Habib Ur Rehman Read Time: 4 min

International Business Machines Corporation (NYSE:IBM) lost 25.2% on July 14, erasing approximately $68 billion in market value in the stock's worst session on record. At first glance, the sell-off appears contradictory amid the still-robust AI investment cycle. Gartner expects worldwide AI spending to jump 47% to $2.59 trillion in 2026, with $1.43 trillion allocated to infrastructure. IBM did not crash due to any weakening in AI spending, but because that spending shifted toward hardware just as the company missed mainframe targets and failed to close major deals.

The AI Boom Hit the Wrong Side of IBM's Portfolio

IBM revealed the damage eight days before its scheduled earnings report. Preliminary second-quarter revenue rose just 1% to $17.2 billion, approximately $660 million below the LSEG consensus, while adjusted earnings of $2.93 per share missed the $3.02 estimate. Software growth slowed to 5%, Consulting was flat, and Infrastructure fell 7%. However, the numbers alone did not appear severe enough to wipe out a quarter of IBM's value; it was the unscheduled warning and CEO Arvind Krishna's blunt admission that the company had "faltered."

In his July 14 investor letter, Krishna stated that clients spent the final weeks of June rushing to secure supply-constrained servers, storage, and memory before expected price increases. Cybersecurity concerns also competed for budgets and attention. IBM did capture some of that spending: Distributed Infrastructure revenue jumped 37%. But that strength was overwhelmed by weakness elsewhere. Several large deals slipped beyond the quarter, while IBM Z and its related Transaction Processing software missed expectations. In summary, the AI boom did not entirely miss IBM; rather, the money hit the wrong side of its portfolio at the wrong time.

That timing affected a stock priced for reinvention. Reuters Breakingviews noted that IBM has spent more than $50 billion since 2018 on acquiring Red Hat, HashiCorp, and Confluent. Before IBM released the preliminary figures, its shares traded at roughly 26 times forward earnings after more than doubling in five years. Its generative AI book had reached $12.5 billion, while first-quarter software revenue grew 11%. The second quarter's 5% growth therefore challenged the premise supporting IBM's rerating. Oppenheimer, which had modeled 12% software growth, responded by downgrading the stock and abandoning its $350 target.

IBM Is a Hold, Not a Buy Yet

Smart-money positioning was mixed before the crash. At the end of the first quarter, 59 hedge funds in Insider Monkey's database owned IBM, down from 64 in the previous quarter, although their combined holdings rose 27% to 6.76 million shares. AQR more than doubled its stake, D.E. Shaw increased its position by 80%, and Millennium cut its common-stock holding by 54%. Meanwhile, short interest rose 14.5% by June 30 to 34.28 million shares, or 3.66% of the float. Skepticism was building, but IBM was not a crowded short. Both datasets predate the July 14 collapse.

This makes IBM a hold, not a stock to buy, simply because it fell 25%. Red Hat grew 11%, AI lifted Consulting signings, and the z17 cycle remains alive, so selling after the collapse appears late. However, the miss also broke the software-growth case supporting IBM's premium. New investors should wait until the slipped deals actually close and software growth recovers from 5%. If both happen without a cut to full-year targets, the selloff was excessive. If not, IBM did not become cheap on July 14; the market simply removed a valuation premium the business no longer deserved.

While we acknowledge the risk and potential of IBM as an investment, our conviction lies in the belief that some other AI stocks hold greater promise for delivering higher returns within a shorter time frame.

READ NEXT: 33 Stocks That Should Double in 3 Years