New York, Feb 25, 2026, 12:28 EST — Regular session
- IBM stock finished higher Wednesday, shaking off earlier volatility sparked by AI-linked concerns about its legacy code business.
- UBS lifted its rating on IBM to Neutral from Sell, while sticking with the $236 price target.
- Investors are questioning if AI coding tools might put a dent in IBM’s demand for services and consulting tied to its mainframes.
Shares of International Business Machines Corp (IBM.N) climbed roughly 4% to $238.74 by midday Wednesday, pushing further off lows from this week’s earlier tumble. Trading has ranged between $231.22 and $239.55, still a long way from the 52-week peak at $324.90. 1
AI-fueled worries over COBOL, that aging programming language still crucial for heavy-duty transaction systems, sparked the latest move. A blog post from start-up Anthropic triggered concern that AI-powered code tools might threaten IBM’s mainframe and consulting business. The result: shares dropped roughly 13% on Monday, Feb. 23, marking the company’s sharpest one-day loss since October 2000. 2
That’s significant: IBM’s Z mainframes—those massive machines powering the guts of operations at banks, airlines, and various government agencies—remain the foundation for much of the company’s steady software and services income. Should AI push modernization into something more standardized and lower-cost, watch for the impact to hit quickly. Pricing would take the first punch, followed closely by a drop in billable consulting hours.
Anthropic said Feb. 23 that Claude Code and similar tools are streamlining the heavy lifting in COBOL modernization, automating exploration and analysis that would otherwise eat up much of the timeline. According to the company, AI lets teams overhaul a COBOL codebase “in quarters instead of years.” 3
IBM wasn’t having it. Software chief Rob Thomas fired off a post on the company’s newsroom, saying, “translating code is one thing” but “modernizing a platform is something else entirely.” His argument: customers aren’t paying for the coding language—they’re buying resilience and security on the platform. 4
Some analysts said markets ran too far with the news. Evercore ISI’s Amit Daryanani described the selloff as “unwarranted.” Jefferies’ Brent Thill, for his part, pointed out that modernization isn’t simply code translation—there’s still integration and change management to deal with. 5
UBS bumped IBM up to Neutral from Sell on Wednesday, holding firm on its $236 price target. The bank pointed to a now more balanced risk-reward following the stock’s 22% drop this year. UBS also mapped out a big spread for possible outcomes: a bull case at $312, a bear case down at $134—reflecting what could happen if AI coding tools start eating into IBM’s software and infrastructure sales, according to its note. 6
IBM has landed a Defense Commissary Agency deal worth up to $112 million over five years, aimed at upgrading electronic shelf labels across commissaries in the U.S. and abroad. “This award underscores our commitment to innovation and operational excellence,” said Susan Wedge, a managing partner for IBM’s U.S. federal market. 7
Still, the stock’s move leaves the main issue wide open: does AI make legacy-code migration less expensive in a way that grows the overall market, or does it drive down costs at the expense of services and consulting margins? If clients demand price cuts or start wrapping up projects more quickly than new spending rolls in, it’s IBM’s services businesses that would feel the pain first.
IBM’s on deck for Morgan Stanley’s Technology, Media and Telecom conference March 3, landing ahead of its Q1 earnings on April 22. Investors want details—mainframe demand, AI code tools, any hint at shifts in the economics for modernization projects. 8
