IBM has survived mainframes, the dot-com bust, and the cloud revolution. What it may not survive, at least in its current form, is the corporate world deciding it no longer needs what IBM sells.

On July 14, 2026, IBM shares fell approximately 25% in a single session, the company’s worst one-day performance since at least 1968. The trigger was a preannouncement of preliminary Q2 2026 revenue guidance of roughly $17.2B, a number that landed well below what analysts had expected. Tens of billions in market capitalization evaporated before the closing bell.

What actually happened

The revenue shortfall was not a manufacturing glitch or a supply chain hiccup. IBM blamed a meaningful shift in how its enterprise clients are allocating technology budgets.

Companies are redirecting spending toward AI infrastructure: servers, memory chips, and storage systems built to run large language models and other AI workloads. That is great news for chipmakers and hyperscale cloud providers. For IBM, which generates significant revenue from software licenses and consulting contracts tied to legacy systems, it is a direct threat to the business model.

Back on February 24, 2026, IBM stock already dropped 13.2%, its worst single-session performance since October 2000. That selloff was partly triggered by the emergence of Anthropic’s Claude Code AI tool, which demonstrated the ability to automate COBOL upgrades on IBM’s legacy systems. COBOL, for the uninitiated, is a programming language that dates to the 1950s and still quietly powers a large portion of the world’s banking and government infrastructure. IBM has long been one of the primary custodians of that infrastructure. The implication of an AI tool that can modernize COBOL code automatically is that enterprises may no longer need IBM’s consultants to do it for them.

IBM’s blockchain bet and what it means now

In October 2025, IBM launched its Digital Asset Haven platform in partnership with Dfns, a digital asset security firm. The platform was built to give institutions a secure way to manage digital assets across more than 40 different blockchains.

The platform targets tokenization, compliance workflows, and institutional custody, which are exactly the services that traditional financial institutions are beginning to demand as they move assets on-chain. IBM’s pitch is that it can bring enterprise-grade security and compliance infrastructure to a space that has historically struggled with both.

The timing is awkward, though. IBM is trying to sell a forward-looking blockchain and AI narrative to investors at the same moment its core business is visibly shrinking. The Digital Asset Haven platform is not generating the kind of revenue that offsets a $17.2B quarter coming in below expectations.