A week ago IBM stock (NYSE: IBM) lost about a quarter of its value in a single trading day, after CEO Arvind Krishna published preliminary results in an unusual step. We broke down the event in real time in a separate piece. This evening came the second act: the full report, with the breakdown that was not included in the early announcement - and also with a guidance update that was not known.
About the Company: A Three-Layer Annuity Machine
IBM is not a technology company that sells a product and moves on. It is built as a recurring-revenue machine in three layers, each leaning on the one beneath it.
The mainframe is the base - enterprise supercomputers that sit in banks, insurance companies, airlines and government agencies, and run the most critical transactions. The real asset is not the hardware but the switching cost: an organization that wrote millions of lines of code on the system over decades does not replace it without a multi-year project of enormous risk.
Software is the middle layer, and at its center Red Hat - the infrastructure layer that allows the same software to run in both public cloud and a private data center. This is the fastest-growing and most profitable layer.
Consulting is the top layer: tens of thousands of consultants who implement systems in organizations. Profitability is lower, but it puts IBM in the room where decisions are made.
About the Report: Exactly What Was Reported
The overall picture. Revenue of $17.2 billion, 1% growth. At the GAAP level, EPS fell 2% to $2.27, and pre-tax profit fell 5%. In operating (non-GAAP) terms the picture is reversed: EPS rose 5% to $2.93, and the pre-tax margin expanded 0.3 percentage point to 19.2%. The gap between the two measures stems mainly from one-time items and amortization - and therefore it is important to look at both.
Software: $7.8 billion, 5% growth. Within it the breakdown reveals a split picture: Red Hat grew 11%, the data segment jumped 19%, automation rose 4% - but transaction processing fell 8%. In other words, the new engines are working while the legacy retreats.
Consulting: $5.3 billion, unchanged (up 1% in constant currency). Standing in place.
Infrastructure: $3.8 billion, down 7% - and here is the sharpest number in the report. Within the segment, IBM Z sales, the mainframe line, plunged 42% - while distributed infrastructure jumped 37%. Two entirely opposite numbers in the same segment.
The explanation for the mainframe drop is cyclical at its base: IBM Z sales move in waves around new generation launches, and a weak quarter at a late stage of a cycle is not the same as a structural break. At the same time, the company notes that Power and storage grew at a record pace, with an order backlog approaching $500 million - meaning demand is flowing to a different part of the same segment.
Cash Flow and Guidance
Cash flow. In the quarter the company generated $2.6 billion from operations - a rise of $0.9 billion. But free cash flow stood at $2.5 billion, a drop of $0.3 billion versus the comparable quarter. In the first half operating cash flow accumulated to $7.8 billion, up $1.7 billion. $1.6 billion in dividends were returned to shareholders in the quarter.
And the guidance - here is the material update. IBM now expects constant-currency revenue growth of 4% to 5% for the full year, versus a previous expectation of growth of over 5%. This is a guidance cut. On the other hand, the free-cash-flow forecast remained in place: growth of about a billion dollars versus last year, and the company even notes that it now expects a larger pre-tax margin expansion. CFO James Kavanaugh phrased it directly: the company "faced revenue headwinds late in the second quarter."
And looking ahead, management emphasized two moves: an investment of over $10 billion in quantum computing over the next five years, with a letter of intent with the U.S. Commerce Department to establish a dedicated quantum wafer plant and a target to stand up a large-scale fault-tolerant quantum computer by 2029; and the launch of Lightwell, an open-source code-security tool that per the company delivered more than 7,500 security fixes in its first two weeks.
The Bull Thesis
Whoever reads the report positively will point out that the new engines are working: Red Hat at 11%, data at 19%, distributed infrastructure at 37% - all three exactly the areas the company marked as the core of growth. Operating EPS rose 5% and the pre-tax operating margin expanded, meaning the company generates more profit from every dollar of revenue even in a revenue-weak quarter.
The free-cash-flow forecast - the basis for the dividend - was not changed, and management is even upgrading margin expectations. The mainframe drop is of a familiar cyclical nature, and a backlog of about half a billion dollars in Power and storage points to demand that keeps flowing in. And finally, after the sharp fall the stock enters the report with low expectations, so the bar for a positive surprise in the coming quarters is lower.
The Bear Thesis
Whoever reads it critically will point out that the guidance was cut - and that is the new and significant number in the report. A company that lowered its growth target mid-year, after already publishing an early warning, signals that the weakness is not merely a one-off.
Revenue growing 1% is slow by any measure. Consulting, which is about a third of revenue, is standing in place. Transaction processing in software is falling 8%. At the GAAP level, EPS, pre-tax profit and gross margin - all fell. And quarterly free cash flow fell $0.3 billion even though operating cash flow rose, a gap that stems from investments and indicates that cash conversion is becoming more challenging.
Beyond that, the over-$10-billion investment in quantum is a long-term bet that will consume cash for years before it returns - and this in a company already dealing with slow growth. And the explanation for the infrastructure weakness, that customers shifted budgets to servers and memory to get ahead of a shortage, places IBM on the wrong side of the money flow in the AI era - exactly the pattern we described in the hybrid-bonding piece.
The debate in one line
The bulls see growth engines that are working (Red Hat 11%, data 19%, distributed infrastructure 37%), an expanding operating margin, and a cash-flow forecast that has not moved - the basis for the dividend. The bears see cut guidance, revenue growing 1%, consulting that is frozen, a mainframe that plunged 42% and quarterly free cash flow that fell. Both sides cite the same report, and both are accurate.
Summary
This report does not provide a verdict - it sharpens the question. IBM presents a company where the new parts are growing nicely and the old parts are retreating, and the pace of transition between the two is the whole story. This quarter, the retreat was fast enough to force a cut in the revenue forecast - and at the same time, operating profitability and the cash-flow forecast were actually maintained or improved.
What can be stated with certainty: this is no longer a growth story, and also not a collapse story. It is a story of transition - and the metrics that will tell whether it is succeeding are free cash flow, the pace of Red Hat and the data segment, and the cyclical recovery of the mainframe. All three we will be able to read better next quarter.
Sources: IBM's official results report for the second quarter of 2026 as filed with the SEC (Form 8-K, exhibit 99.1, July 22, 2026), including the segment breakdown, cash-flow figures and management statements attached to it. The chart is shown in real time via TradingView.