IBM Lowers Full-Year Revenue Guidance, Can It Still Price Based on Stable Cash Flow?

TL;DR · IBM's stock fell about 25% on July 14 after a warning, and on July 22, it officially lowered its full-year revenue guidance. · The market is divided on whether the budget squeeze from AI infrastructure is a quarterly deferral or a loosening of traditional enterprise IT valuation anchors. · Related entities: IBM, enterprise software, IT service companies, AI servers, storage and memory supply chain.

After a warning on July 14, IBM's stock fell about 25% in a single day, and on July 22, it officially lowered its full-year revenue guidance in its second-quarter report. The trigger for the sell-off was not a few percentage points drop in revenue, but the company's admission that clients had temporarily shifted capital expenditures in the last few weeks of June towards servers, storage, and memory needed for AI.

This statement is very sensitive for investors. In recent years, the market has been willing to give companies like IBM a stable cash flow valuation because mainframes, transaction processing software, and enterprise services do not fluctuate as violently as consumer electronics. However, this time the market discovered that even critical systems like banking and aviation could be temporarily sidelined by clients when AI infrastructure supplies are tight.

IBM CEO Arvind Krishna acknowledged in a July 14 investor letter that the company did not anticipate the extent of the capital expenditure reordering by clients at the end of June and stated that the execution this quarter was lacking. In a conference call, he provided a counter-explanation, stating that much of the demand is "deferred, not disappeared," and that about one-third of the large capital expenditure transactions that had fallen off were completed in the first three weeks of the third quarter.

The market now needs to assess the distance between these two statements. Is this a quarterly mismatch exacerbated by supply chain panic, or an early signal that AI capital expenditures are beginning to compress traditional enterprise IT annuity revenues?

The Plunge Prices a Zero-Sum Budget

IBM's total revenue for the second quarter was about $17.2 billion, a year-on-year increase of 1%. Software revenue was about $7.8 billion, up 5%. Infrastructure revenue was about $3.8 billion, down 7%, with IBM Z mainframe revenue down 42% year-on-year. The company adjusted its full-year constant currency revenue growth expectation for 2026 from "above 5%" to 4%-5%, while maintaining its free cash flow expectation to increase by about $1 billion year-on-year.

Looking solely at the magnitude of the guidance downgrade, it does not seem sufficient to explain the approximately 25% single-day drop. The real trigger for the revaluation is structural. Mainframes are inherently cyclical businesses; the market can tolerate fluctuations as long as they do not harm the higher-quality software cash flows.

The issue this time is that clients are not simply buying one less machine; they are reallocating budgets to secure servers, storage, and memory needed for AI. The so-called capital expenditure reordering means that the money originally planned for system upgrades or software purchases is being temporarily redirected to lock in infrastructure capacity due to concerns about shortages and price increases in AI hardware.

This has transformed IBM's decline from a financial report misstep into a valuation anchor shake-up. If AI infrastructure becomes a priority in enterprise budgets, traditional IT companies that were once seen as stable revenue streams may need to be repriced.

Mainframe Decline Will Affect Software Revenue

Ordinary investors may misunderstand mainframes, thinking they are just outdated hardware. In reality, IBM Z mainframes remain core devices for many banks, airlines, insurance companies, and government systems handling critical transactions, characterized by reliability, security, and high throughput, with high migration costs.

More importantly, mainframes are not sold as standalone hardware. They are typically bundled with transaction processing software, which handles high-frequency critical transactions, as well as long-term maintenance services. These software products have high margins and stable renewals, forming part of the quality of IBM's software business.

Reuters quoted IBM's CFO stating that the mainframe hardware and transaction processing software stack has caused a drag of over 5 percentage points on IBM's growth, whereas the previous expectation was about 1-2 percentage points. This statement should not be simply interpreted as software dragging down performance alone, but it indicates that the hardware cycle has already impacted the rhythm of software and service revenues.

For the market, the danger signal is not that IBM is selling fewer mainframes, but that the hardware cycle, software licensing, and customer budget decisions are all being penetrated by the AI supply chain shock in the same quarter. Revenue streams that were originally priced separately suddenly appear to be more correlated.

This is also why investors are concerned about the pressure spilling over to some software and IT service companies. The market does not believe that all software will immediately decline, but is instead re-examining a question: If clients exceed their AI budgets, which traditional system upgrades will be postponed?

The Limits of Krishna and Idlet's Counterarguments

Krishna's counterargument focuses on rephrasing "demand disappearing" to "timing of transactions being pushed back." He emphasized that there is no evidence of clients migrating away from mainframes, that the z17 project is still strong, and that Red Hat's growth is also improving, with some delayed transactions already completed in early Q3.

This explanation can alleviate the most extreme pessimistic narrative. If delayed orders in Q3 and the second half of the year continue to materialize, the second quarter will appear more like a short-term mismatch caused by tight AI hardware supply, compounded by IBM's own execution not adjusting in time.

CFRA analyst Brooks Idlet also stands in this direction. He believes that IBM's software issues are more likely to reflect the company's own hardware problems rather than an overall trend in the software industry. The effect of this viewpoint is to attempt to sever the spillover chain between "IBM's plunge" and "the software industry's budget crisis."

However, the counterarguments do not completely eliminate the risks. IBM has already confirmed that clients shifted budgets towards AI infrastructure at the end of the quarter. The zero-sum relationship of budgets has at least occurred among some large clients and within certain time windows. The only difference is whether it will happen again.

If AI servers, memory, and storage supplies remain tight, clients may not just postpone traditional system projects once. A single deferral can be recovered, but multiple deferrals can change the annual growth rhythm and may lower the market's willingness to assign stable revenue valuation multiples.

Order Fulfillment Determines Whether Valuation Can Be Repaired

This time, IBM's plunge cannot simply be written off as "AI killing mainframes." Current evidence supports a narrower judgment: Tight AI supply in Q2 amplified IBM's mainframe cycle and execution issues, and led the market to reassess the stability of traditional enterprise IT revenues.

The verification point lies in Q3 and the second half of the year. If the delayed demand that Krishna mentioned continues to convert, and the drag from mainframe-related hardware and transaction processing software narrows, annual growth can still fall within the new guidance range, and the market will attribute this shock more to timing issues. At that time, the approximately 25% drop on July 14 may have included excessive panic.

Conversely, if the fulfillment of delayed orders does not go smoothly, or if more enterprise software and IT service companies mention that clients are postponing traditional projects for AI infrastructure in their earnings reports, Idlet's explanation of "IBM's individual hardware issues" will be weakened. At that point, the market will not just be trading on IBM's misstep, but on the ongoing pressure of AI capital expenditures on traditional enterprise IT budgets.