Mythos, Quantum, Slowing Sales: What Are the Main Threats to IBM?
Following the release of its preliminary quarterly results, several analysts immediately lowered their forecasts for IBM. Photo: Carson Masterson / Unsplash.com
On the evening of June 22, IBM will release its quarterly earnings report. The release of preliminary results a week ago led to the company’s biggest stock plunge in nearly 60 years and dragged down shares of software companies. But the market has already recovered from that shock. What should investors really be keeping a close eye on?
The market is reassessing the risks
On July 14, IBM’s stock price fell 25% in a single day after the company released preliminary results for the second quarter of 2026. The report indicated that its revenue and adjusted earnings per share fell short of analysts’ expectations ($17.2 billion and $2.93 versus $17.86 billion and $3).
The company's CEO, Arvind Krishna, called the results "disappointing." The main reason was weaker sales of IBM Z mainframes (enterprise servers) and related software. The company’s customers began allocating their budgets to the purchase of servers, data storage systems, and memory, and spent less on IBM products. “We were unable to adapt quickly enough to the changing situation and failed to close many major deals within the expected timeframes,” Krishna acknowledged.
However, HSBC believes that the market has overestimated the extent of the problems with mainframe sales—customers may simply have preferred to sign shorter-term contracts.
According to HSBC, IBM faces a more serious threat: new AI models, such as Mythos from Anthropic. Part of the company’s software business relies heavily on open-source software. It likely suffered due to concerns that such AI models could allow attackers to analyze open-source code more quickly and find vulnerabilities, according to the bank’s July 16 note.
Evgeny Shatov, an investment manager and partner at the investment firm Capital Lab, believes that the Mythos risk is quite serious. The model poses risks primarily for software developers whose businesses rely heavily on open source, including MongoDB, Elastic, GitLab, and JFrog. At the same time, he believes it could stimulate demand for cybersecurity solutions and software supply chain protection.
AI reduces the cost and time required to find vulnerabilities in open-source code. This could increase security spending at IBM and Red Hat (an IBM division specializing in enterprise open-source software—Ed.) and slow down customers’ purchasing decisions. But there is also another side to this: the more dangerous free open source becomes, the greater the value of IBM’s commercially supported and updated solutions.
To understand which side IBM will end up on in this story, investors need to keep an eye on the progress of Lightwell—the company’s $5 billion project to protect against and fix vulnerabilities in open-source components—and its actual revenue stream, he believes.
The ATLAS Capital Research Department also believes that Red Hat’s sales and revenue figures should be monitored. According to the analysts, these figures are a “direct indicator that Mythos does not yet pose a serious threat” to IBM.
The market is only just beginning to assess who will be a net buyer and who will be a provider of security infrastructure. IBM’s current price already factors in a premium for uncertainty. However, specific long-term losses or gains may not be fully reflected, as there is no data on Lightwell’s revenue or the pace of adoption.
Not Just AI
JPMorgan has identified five additional key risks for IBM.
Two of them are common to the market: a possible slowdown in the global economy and a slower adoption of AI.
He sees another opportunity in IBM’s investments in quantum computing. In June of this year, the company announced a five-year plan to invest $10 billion in this field, with the goal of creating the world’s first quantum computer capable of solving complex practical problems by 2029. JPMorgan notes that these investments are increasing the company’s capital expenditures and could put pressure on its free cash flow. Furthermore, breakthroughs by competitors also working on quantum computing—including Microsoft and Google—could lead to a decline in its stock price.
Another risk is a potential slowdown in sales of Red Hat’s enterprise operating systems. And finally, the last risk is slower integration of IBM’s recently acquired companies, primarily Confluent (for $11 billion). If integration or cross-selling progresses more slowly than expected, the impact of the deals may fall short of the projected synergies of approximately $500 million per year.
What other metrics should you monitor?
IBM’s stock never fully recovered from the crash. In the week following the release of its preliminary quarterly results, at least eight Wall Street analyst teams lowered their price targets for IBM, and one bank — Oppenheimer — withdrew its target price (literally the very next day after raising it from $320 to $350) and downgraded its rating from “outperform” to “neutral.”
Currently, 16 out of 28 analysts recommend buying IBM stock, 10 recommend holding, and two recommend selling. The average price target is $264.1 per share, which implies a 25.5% increase from current levels.
The market priced in not only a weak quarter but also three more serious risks: reduced business predictability, the Software segment’s dependence on IBM Z, and management’s inability to quickly adapt sales to changes in the company’s customers’ budgets.
Going forward, the decisive factors for investors will not only be the already known revenue and earnings per share figures, but also the free cash flow forecast. According to analysts at ATLAS Capital, it is precisely this metric that will reveal whether the current difficulties are temporary or reflect a deterioration in the business model.
The revenue trends in the software business and Red Hat's financial metrics are also important. The company must also demonstrate that major deals were indeed deferred rather than completely lost.
This article was AI-translated and verified by a human editor
