Markets IBM JUL 19, 2026

IBM crashes 25% in worst day since 1916 listing as AI capex shift guts software line

Krishna blames late-June client pivot to servers, storage, and memory; Salesforce, ServiceNow, and Alphabet drag in the sector rout.

BY HENLEY MARRAST · MARKETS DESK

IBM shares fell 25% on Tuesday, the worst single-day drop since the company listed on the NYSE in 1916, eclipsing the 23.7% Black Monday drawdown of Oct. 19, 1987 and every session in the modern trading record that began in 1968. The proximate cause was a preannouncement letter from CEO Arvind Krishna warning that preliminary Q2 revenue would come in at $17.2 billion against a FactSet consensus of $17.86 billion, with adjusted EPS of $2.93 versus the $3.01 Street number (Reuters had $3.02). Growth of roughly 1% is IBM’s weakest in more than a year.

The explanation is where it gets interesting. Krishna told investors the software business “faltered” in keeping pace with a spending shift out of software and into data-center hardware. “In the last few weeks of June, we saw clients shift their quarterly capex spend toward servers, storage, and memory purchases to secure supply-constrained infrastructure ahead of expected price increases,” he wrote.

Read that carefully. IBM isn’t losing to a competitor. It’s losing to a category, in real time, inside a single quarter. Enterprise budgets that had been earmarked for seat licenses got redirected mid-June toward physical AI supply.

The tape confirmed the rotation. Salesforce dropped 2% Tuesday and ServiceNow nearly 6%, even as Salesforce booked a 4.6% weekly gain and sits down 35.5% year-to-date. CNBC has taken to calling it “the great AI rotation.” The hardware side wasn’t spared either when supply anxieties flipped the other way: SK Hynix’s U.S.-listed debut fell 9% on July 13, and Sandisk, Intel, and AMD opened the same week down 12%, 6%, and 4%.

Then Alphabet slid 4.5% intraday Thursday and closed roughly 6% lower over two sessions on Bloomberg reporting that Gemini 3.5 Pro, announced at Google I/O in May, is months behind schedule. A company spokesperson said Google was “shipping quickly across a wide range of models while keeping them highly cost-effective.” Bloomberg separately noted Chinese chip-supply-chain names posting preliminary profit gains topping 1,000% have failed to revive their rally, the upside already priced in.

The doom loop here isn’t AI skepticism. It’s the opposite: capital so convinced of the buildout that it’s willing to strand a century-old software franchise in a single afternoon to fund it.

Sources

About the author

Henley Marrast

MARKETS DESK

Henley Marrast covers AI-equity flow, accelerator demand, and earnings prints for AI Sheet Report. She leads coverage of the public AI complex from the New York markets desk, with a focus on the daily tape and quarterly results. She has been writing about technology markets for several years.