IBM shares climbed about 5.2% by 10:50 a.m. ET on Tuesday, bouncing back after a heavy fall on Monday. The drop happened after Anthropic said its Claude Code tool can help modernise old COBOL systems, which made investors worry that AI could hurt IBM’s mainframe business.
But analysts say those fears are likely overdone. Companies have had many years to move away from mainframes, but still they use them as they are extremely reliable to handle huge volumes of transactions and rarely fail. For banks and large businesses, that kind of stability is hard to replace.
Evercore analyst Amit Daryanani told Investing.com that customers continue to use mainframes due to advantages including “100% uptime, better cost efficiency at scale, on-premise AI inferencing capabilities, quantum-safe encryption, and regulatory considerations for sensitive industries.”
IBM is already using AI to modernise COBOL
Another big reason for the rebound was that IBM is not being left behind by AI, it is already using it. Jefferies analyst Brent Thill reminded investors that IBM has been building its own AI-powered tools to handle COBOL updates. “Watsonx Code Assistant for Z has already been in production for over two years and removes the traditional burden of mainframe modernization by using GenAI to refactor COBOL into Java, explain production code, and modernize applications while preserving critical logic,” Thill wrote.
“Embedding these capabilities directly into the Z (mainframe) platform gives IBM a structural advantage over horizontal code assistants, which may be powerful but lack native access to mainframe data, tooling, and operational context.” In other words, IBM is not watching disruption happen but it is building it into its own systems.
Strong demand for IBM’s new mainframe systems
There is also actual business momentum behind the stock. IBM’s latest z17 system is reportedly performing better than the previous z16 cycle at the same stage. System-Z hardware revenue has grown sharply, showing customers are still investing in these machines. That suggests companies are not rushing to abandon IBM’s technology and they are still buying it.
Monday’s IBM crash
Monday’s 13% crash was IBM’s worst single-day drop in more than 25 years. The stock had already fallen significantly from its recent highs. Some analysts believe the market reaction was emotional rather than based on fundamentals. After the drop, IBM trades at around 18 times expected 2026 earnings and about 16 times this year’s free cash flow — levels some see as reasonable for a company with steady enterprise demand. As a result, investors stepped in to buy the dip.
IBM continues to expand its AI ecosystem. The company had previously partnered with Anthropic to integrate Claude into parts of its software portfolio.
It also announced a collaboration with Deepgram to add speech-to-text and voice capabilities to its watsonx Orchestrate platform. Deepgram CEO Scott Stephenson said, “Voice is rapidly becoming the default interface between humans and technology, and enterprise deployments require a real-time platform that is accurate, low latency, and reliable at scale.”
IBM executive Nick Holda added, “Our watsonx Orchestrate integration powered by Deepgram APIs introduces new speech recognition and transcription capabilities to IBM clients, refining and modernizing their operations.” These updates show IBM is pushing deeper into AI tools for businesses and not retreating.
The initial sell-off was due to worries that AI could replace parts of IBM’s legacy business. But analysts argued that mainframes remain deeply embedded in global finance and transaction systems. Once investors heard that IBM is both defending and upgrading its core systems with AI and still seeing strong product demand confidence improved.
Disclaimer: This article provides factual analysis only and is not, and should not be construed as, an offer, solicitation, or recommendation to buy or sell securities. Investors must conduct their own independent due diligence and seek advice from a registered financial advisor in the respective jurisdiction.