IBM Shares Slide as Anthropic's Claude Code Raises Questions About Mainframe Demand
Published on 11/17/2026 at 18:11 | Editorial boerse-global.deIBM stock fell 4.3% on Wednesday, caught in a broader technology selloff driven by fresh U.S. restrictions on China's access to advanced AI chips. Investors were also holding back ahead of Nvidia's quarterly results and Anthropic's planned initial public offering, according to Reuters. But beneath the sector-wide caution lies a more company-specific debate: what increasingly capable AI coding tools mean for IBM's mainframe and consulting businesses.
Claude Code Enters the COBOL Conversation
Anthropic said in an update Tuesday that Claude Code can assist with modernizing COBOL, the decades-old language still underpinning many mainframe systems. Media coverage of the announcement stirred concerns among investors about potential effects on demand for IBM's mainframe and consulting services. The reporting links Anthropic's move to questions about how much need companies will have for those IBM offerings going forward.
That tension — new applications opening up possibilities while calling established workflows into question — sits at the heart of how the market is reading IBM during the current tech rout. General caution toward AI stocks has now fused with a concrete discussion about the company's core franchises.
IBM, for its part, is pressing ahead with enterprise AI tools of its own. On November 10, the company and Microsoft unveiled the Imagine Enterprise platform, designed to help businesses build, run and govern AI agents for industry-specific processes.
Should investors sell immediately? Or is it worth buying IBM?
A Study That Shifts the Focus to Execution
How difficult that organizational work remains is underscored by a study published Wednesday by the IBM Institute for Business Value, conducted with Oxford Economics. Only 18% of surveyed AI decision-makers say they have full control over decisions about AI investments. The survey covered 1,667 executives and was fielded from April to June 2026, describing real-world corporate experience rather than expectations about what technology might someday make possible.
For anyone evaluating IBM, that distinction matters. Progress in AI or quantum computing can open technological vistas, but a durable investment case requires evidence that companies are actually extracting economic value from those capabilities. The study offers benchmarks, not a revenue forecast for IBM.
The key metric is the share of AI projects delivering measurable business results. Among the leadership group IBM identifies as most effective, that figure stands at 63%, versus 43% for other respondents. What counts is not the number of projects launched but their economic return. The survey shows correlations, not causation — a better success rate does not prove that any single management practice automatically produces better outcomes — yet it remains a useful yardstick.
The gap persists at the handoff from experimentation to production: 49% of initiatives make that leap in the top group, compared with 33% elsewhere.
Discipline as the Positive Case
The optimistic scenario starts with more rigorous investment decisions. Successful initiatives scale within 8 months in the high-performing group, while the rest take 13.5 months. Shorter paths from pilot to broad deployment could help companies identify viable projects faster, laying a firmer foundation for further spending — a more convincing signal for IBM than an assumption of ever-rising budgets with no verifiable payoff.
Willingness to pull the plug stands out too: 96% of the top group halt or cut funding for unsuccessful AI projects, against 46% of the comparison group. For investors, a stricter selection process need not undermine the long-term AI story. If money flows instead toward effective applications, tougher screening could strengthen the technology's economic footprint.
IBM at a turning point? This analysis reveals what investors need to know now.
Where Progress Gets Stuck
The bearish scenario lives in the gap between technical possibility and organizational capability. According to IBM, 75% of respondents say governance, risk or compliance concerns delay AI actions, and just 29% say their organization responds quickly to insights about AI performance. Such delays are not mere bureaucracy: when AI systems act autonomously, accountability and control become economically relevant. Unclear rules could slow adoption and limit the value of applications already tested.
Already, 48% of respondents report that AI agents execute actions across workflows, giving added weight to the question of who oversees such systems and can intervene when problems arise. Equating technical progress directly with incremental business revenue would therefore be risky. If decision rights remain fragmented, even compelling applications may linger in testing. The study documents this implementation hurdle but does not quantify any financial impact on IBM.
Nvidia Numbers and a Quantum Testbed
Near-term, the next concrete marker for the AI market landscape arrives with Nvidia's quarterly results, due after the close. They are no direct evidence of IBM's business trajectory, but they could shape the expectations investors apply to other technology bets. Separately, IBM is working with Nvidia and Fermilab on a modular quantum test platform. It will initially emulate communication between quantum modules, with physical connections to follow later — a research prospect, not a completed commercial breakthrough. Technical advances and proven business results, in other words, deserve separate yardsticks.
Ad
IBM Stock: New Analysis - 17 November
Fresh IBM information released. What's the impact for investors? Our latest independent report examines recent figures and market trends.
