Key Takeaways
International Business Machines Corp. (IBM) shares took a significant hit recently, plunging 13% on Monday, February 23, 2026, marking its steepest single-day percentage loss since October 2000. This dramatic sell-off was triggered by AI startup Anthropic PBC, which announced its Claude Code tool can help modernize COBOL, a decades-old programming language central to many critical enterprise systems. The market's immediate reaction was a clear signal of fear that AI could rapidly erode a significant portion of IBM's legacy business.
The concern stems from COBOL's pervasive, yet often overlooked, role in global infrastructure. This language, developed in the late 1950s, still powers an estimated 95% of ATM transactions in the U.S. and underpins core systems in banking, aviation, and government. Maintaining these systems is expensive, largely due to a shrinking pool of COBOL-proficient programmers. Anthropic's claim that Claude Code can automate the "exploration and analysis phases" of COBOL modernization, dramatically cutting costs and time, directly challenged IBM's long-standing revenue streams from managing and migrating these legacy systems.
IBM shares, which closed at $223.35 on Monday, saw a rebound today, trading at $233.22, up +4.42%. However, the stock remains down 27% for February, putting it on track for its largest one-month slide since at least 1968. This volatility underscores investor anxiety about how quickly generative AI tools might disrupt established enterprise software and consulting models, particularly for companies like IBM that have historically profited from the complexity of legacy IT.
The market's knee-jerk reaction highlights a broader trend: investors are increasingly sensitive to any AI-related development that could diminish demand for legacy products or undermine traditional licensing, maintenance, and consulting revenue streams. The ability to "vibe code"—using AI to write software—is seen by some as a direct threat, enabling enterprises to create their own applications and potentially reducing reliance on established vendors. For IBM, whose mainframe business is deeply intertwined with COBOL-based systems, this perception translated into a swift and severe market repricing.
Anthropic's claims about Claude Code's ability to streamline COBOL modernization are compelling, suggesting that AI can analyze massive COBOL codebases, flag risks, and drastically reduce the effort and cost typically associated with such projects. This capability, if fully realized, could indeed disrupt the traditional model where "armies of consultants spending years mapping workflows" were required. The market interpreted this as a direct attack on IBM's high-margin "discovery" phase of consulting contracts and its infrastructure lock-in.
However, IBM has been quick to defend its position, emphasizing that the "value IBM mainframe delivers has nothing to do with COBOL." IBM Senior Vice President Rob Thomas stated that whether an application is written in COBOL, Java, or any other language, the mainframe platform provides the same guarantees of performance and security. This argument suggests that the underlying platform's reliability, rather than the specific programming language, is the true source of its value to mission-critical operations in finance, government, and other highly regulated industries.
Moreover, the idea of AI-assisted COBOL modernization is not new, nor is it exclusive to Anthropic. IBM itself debuted a tool in 2023 that uses AI to work with COBOL, including updating it into Java. Its watsonx Code Assistant for Z solution, which uses a 20-billion parameter model, is designed to translate COBOL to Java and has already shown significant improvements in developer productivity. CEO Arvind Krishna noted in July 2025 that IBM's AI coding assistant for mainframes has "very wide adoption," primarily for understanding COBOL code and deciding what to modernize.
Analysts like Amit Daryanani of Evercore ISI point out that clients have "already had the option to migrate from the mainframe, yet they are sticking with the platform." This suggests that the perceived "erosion of complexity" might be overblown, as enterprises prioritize the mainframe's stability, security, and regulatory compliance over potentially cheaper, but riskier, migration alternatives. While AI can undoubtedly accelerate parts of the modernization process, the deep integration and mission-critical nature of these systems often necessitate incremental changes rather than wholesale shifts, a reality that favors IBM's established expertise.
IBM's strategy for the mainframe is far from static; it's deeply integrated with its broader AI vision. The company views AI not as a threat to its mainframe business, but as a powerful enhancement. IBM's latest mainframe, the IBM z17, is designed for enterprise-scale AI, featuring built-in AI capabilities powered by the Telum II processor and the new Spyre AI accelerator. This hardware enables real-time AI inferencing directly on the system, allowing clients to leverage their vast amounts of mainframe data for immediate insights without moving it off-platform.
The company is actively infusing AI into transactional workloads to obtain real-time insights and building intelligent infrastructure using AI for mainframe operating systems. This includes tools like watsonx Code Assistant for Z and watsonx Assistant for Z, which deliver intelligent support for developers and IT teams. These AI assistants help bridge the growing knowledge gap as veteran mainframe professionals retire, providing code explanation, automated refactoring, and optimization advice, making it easier for new developers to maintain and update COBOL applications.
IBM's approach emphasizes a hybrid cloud strategy, recognizing that many organizations find it optimal to use the cloud for general knowledge functions while maintaining on-premises deployments for secure, mission-critical automation tasks. This aligns perfectly with the mainframe's strengths in security and data residency, particularly for highly sensitive data in finance, healthcare, and government. An IBM Institute for Business Value report indicates that 79% of IT executives believe mainframes are essential for enabling AI-driven innovation.
Ultimately, IBM's strategy is about bringing AI to where enterprise work already happens, maximizing the value of existing mainframe investments while introducing new capabilities. This includes using AI to predict upcoming workloads, proactively prepare resources, detect system operation anomalies, and assist in predicting outages. The company argues that AI strengthens the mainframe case by compressing timelines for code refactoring, DevOps modernization, and knowledge preservation, all of which are on-platform opportunities that AI accelerates.
Despite the recent stock plunge, IBM's financial fundamentals paint a picture of a resilient, cash-generating enterprise. The company reported EPS of $4.52 in its last earnings call on January 28, 2026, beating estimates, with revenue reaching $19.7 billion. For the trailing twelve months (TTM), IBM's revenue stood at $72.11 billion, with a net income of $11.31 billion, translating to an EPS of $11.31. These figures demonstrate a company that continues to deliver substantial financial performance.
Looking at valuation, IBM trades at a P/E ratio of 20.62, a P/S of 3.23, and a P/FCF of 17.75. While these metrics reflect a mature technology company, they are not indicative of a business on the brink of collapse. The company's market capitalization stands at $217.99 billion, with an Enterprise Value (EV) of $271.51 billion. Its gross margin is a healthy 58.8%, with an operating margin of 17.3% and a net margin of 15.7%, showcasing strong profitability from its operations.
IBM's commitment to shareholder returns is evident in its dividend, currently at $1.68 per share quarterly, offering a yield of 2.9%. The payout ratio of 59.0% suggests the dividend is well-covered by earnings. Furthermore, the company has demonstrated solid growth in key areas: TTM net income grew by 75.9% year-over-year (FY2025 YoY), and EPS grew by 73.7%. While operating cash flow and free cash flow saw slight declines of -1.9% and -1.6% respectively, the overall picture is one of a company effectively managing its financial performance.
Wall Street analysts currently have a "Hold" consensus rating on IBM, with 23 out of 49 analysts recommending a hold. The consensus price target is $349.50, with a median of $347.50 and a high of $380.00, significantly above the current price of $233.22. This suggests that many analysts believe the recent sell-off may be an overreaction, and there is substantial upside potential if IBM can effectively navigate the AI landscape and reassure investors about its long-term strategy. The company's deep integration in highly regulated industries, where stability and security are paramount, continues to provide a strong economic moat.
For investors, the Anthropic news presents a classic bull versus bear scenario for IBM. The bear case hinges on the fear that AI-driven tools like Claude Code will fundamentally erode IBM's lucrative mainframe and consulting businesses. If AI can truly make COBOL migration "press-a-button" simple, the high-margin "discovery" phase of consulting contracts could vanish, and the economic "gravity" that keeps banks and governments tied to mainframes might weaken as exit costs plummet. This perspective sees IBM as a legacy company struggling to adapt to a rapidly evolving AI ecosystem, risking diminished demand for its enterprise software platforms and AI integration services.
The bull case, however, argues that the market's reaction was an overcorrection. Proponents point to IBM's own significant investments in AI for mainframes, such as watsonx Code Assistant for Z, which has been in development since 2023. They emphasize that mainframe customers prioritize reliability, security, and regulatory compliance, factors that AI tools alone cannot fully replace. The tight coupling of software and hardware on IBM Z systems offers performance and security guarantees that are difficult to replicate by simply moving code to other platforms. IBM's deep relationships with enterprise clients in critical sectors like finance and government provide a strong, sticky customer base.
Consider these points for a balanced view:
The long-term outlook for IBM will depend on its ability to effectively communicate and execute its AI-enhanced mainframe and hybrid cloud strategy. Investors should watch for continued adoption of watsonx Code Assistant for Z, further integration of AI into the IBM Z platform, and sustained revenue from its enterprise clients.
The recent market volatility surrounding IBM and Anthropic's COBOL AI tool underscores a critical juncture for Big Blue. While the immediate reaction was sharp, it's crucial to differentiate between perceived threats and the complex realities of enterprise IT modernization. IBM's core mainframe business, with its unparalleled reliability and security, remains indispensable for many of the world's largest and most critical organizations.
IBM's proactive integration of AI into its mainframe offerings, coupled with its robust hybrid cloud strategy, positions the company to evolve rather than be disrupted. The challenge now is to effectively articulate this vision to investors, demonstrating that AI is a tool for enhancement and efficiency within its ecosystem, not an existential threat. For investors, this means looking beyond the headlines and evaluating IBM's strategic execution and its ability to leverage AI to strengthen its long-standing relationships with enterprise clients.
The future of IBM will likely be defined by its success in transforming legacy systems from within, using AI to bridge skill gaps and accelerate modernization on its own terms. This isn't just about protecting old revenue streams; it's about upgrading them and proving that the mainframe, far from being a relic, is a dynamic platform capable of embracing the AI era.
Want deeper research on any stock? Try Kavout Pro for AI-powered analysis, smart signals, and more. Already a member? Add credits to run more research.
