I'll research IBM before writing the preview. Let me gather recent earnings, guidance, news, and analyst expectations.This is a huge story — IBM pre-announced a Q2 miss on July 14 and the stock crashed ~25%. Let me read the pre-announcement details and prior quarter's guidance for full context.I have strong context. Let me pull the stock price reaction and check for the latest analyst/news developments right before the call.I now have a comprehensive picture. Here's the earnings preview.
This is the most important IBM earnings call in years — not because the numbers are unknown, but because IBM already pre-announced a badly-missed quarter on July 14 and the stock crashed ~25% in a single day, the worst one-day drop in the company's 115-year history. The call is now entirely about one thing: can management restore credibility and give investors a full-year guide they can trust?
In a highly unusual move, IBM released selected preliminary Q2 results via a letter from CEO Arvind Krishna on July 14, rather than waiting for the scheduled call. The prelims:
| Metric | Prelim Q2'26 | Y/Y | Street Est. | Result |
|---|---|---|---|---|
| Revenue | $17.2B | +1% | ~$17.86B | Miss |
| Operating (non-GAAP) EPS | $2.93 | +5% | ~$3.01–3.02 | Miss |
| GAAP EPS | $2.27 | –2% | — | — |
| Software revenue | +5% | — | (double-digit expected) | Miss |
| Consulting revenue | flat (+1% cc) | — | — | Soft |
| Infrastructure revenue | –7% | — | (down low-single-digit expected) | Big miss |
| Operating gross margin | 59.4% | –70bps | — | Softer |
| Operating pre-tax margin | 19.2% | +30bps | — | Held up |
| YTD Free cash flow | $4.8B | — | — | On track |
The critical nuance: the shortfall was almost entirely a revenue and mix problem, not a profitability collapse. Operating pre-tax margin actually expanded 30bps and YTD FCF ($4.8B, on ~$7.8B operating cash) is intact. IBM's cost/productivity flywheel is still working; the top line broke.
Krishna's letter attributed the miss to a confluence of factors, admitting outright that "this quarter we faltered" and that IBM "did not adapt and move quickly enough":
The bull/bear crux is #2 and #4: Is this a temporary pause (deals slip, then close; capex normalizes) or a structural shift of enterprise budgets away from IBM's software/infrastructure toward AI hardware? Krishna essentially needs to answer that question convincingly on the call. Bloomberg Intelligence has flagged this may not be a one-off.
The prelim letter was deliberately seeded with positives to counter the narrative: - Red Hat re-accelerated to +11% (up sequentially from 10% in Q1) — directly relevant to the "is software durable in an AI world" debate. - HashiCorp and Confluent both performing strongly — the M&A engine is working. - Distributed Infrastructure had its best quarter on record, +37% (Power + Storage), with ~$500M backlog. Ironically, the same AI-hardware wave hurting the mainframe is helping IBM's Power/Storage business. - z17 remains ~130% program-to-program vs. z16 (IBM's prior best), with 85% of installed MIPS maintaining/growing capacity — management's argument that the Z weakness is timing, not demand destruction. - Continued Consulting GenAI signings momentum. - New initiatives: Lightwell ($5B commitment, an open-source security "clearinghouse" with marquee bank clients) and >$10B quantum investment over 5 years, including the "Anderon" quantum wafer foundry with $1B CHIPS support.
Since the headline numbers are known, the entire event hinges on the outlook. Krishna explicitly said the July 22 call will "discuss our full-year expectations."
#1 — Full-year guidance revision. Coming into the year IBM guided: - Constant-currency revenue growth of 5%+ - Software 10%+ (raised at Q1) - Consulting accelerating to low-to-mid single digits - Infrastructure down low-single-digits - Operating pre-tax margin +~1pt - Free cash flow +~$1B (toward the ~$13.5–14B range)
With Q2 software at only +5%, consulting flat, and infrastructure down 7%, the 5%+ revenue and 10%+ software targets look very hard. Watch whether IBM (a) cuts guidance and by how much, (b) reaffirms FCF even if it cuts revenue, or (c) declines to give a clean guide — the last of which would be the worst outcome for a market already questioning credibility.
#2 — Free cash flow defense. FCF is now the metric the Street values IBM on (management's own framing). If IBM can hold ~$13–14B FCF despite the revenue miss, it supports the dividend and the valuation floor. The dividend looks safe — ~$6B cost against ~$13B trailing FCF — and the yield has jumped to ~3.1% after the drop.
#3 — "Pause vs. structural" evidence. Do the slipped deals show signs of closing in Q3? Is the June capex shift reversing? Any hard data here will move the stock more than the guide itself.
#4 — Software durability. Given the broader "will AI make enterprise software obsolete" fear (which dragged Microsoft, Oracle, Salesforce, Workday down on IBM's warning, and the Starbucks/AI-replacing-software story from July 9), IBM needs to reinforce the Red Hat/data/automation acceleration story.
#5 — Governance/strategic overhang. Reports indicate IBM's board recognizes the stock weakness could invite an activist campaign, and there is public debate about whether IBM should be broken up. Any commentary on portfolio, buybacks, or M&A appetite will be scrutinized. (In Q1, Krishna hinted at a bigger 2H M&A appetite given attractive software valuations.)
IBM ran to a near-record ~$306 on July 7, then collapsed: - July 13 (day before warning): $290 - July 14 (warning): $217 (–25%) - July 21 (day before call): $210.54
That's roughly –31% from the peak in two weeks. The stock had been a crowded "AI infrastructure winner," which amplified the punishment. Mizuho's read is that the ~25% drop prices in a more prolonged slowdown than management has implied — meaning if IBM delivers a credible, only-modestly-cut guide, there is room for a relief bounce; if guidance is vague or cut hard, the de-rating continues.
Sell-side is fracturing: HSBC cut to Reduce (Street-low ~$191 PT), arguing investors are better off in a "synthetic IBM" basket; Susquehanna had initiated Neutral (~$303) just before the crash, flagging both quantum upside and the risk that AI-assisted COBOL modernization erodes IBM's high-margin mainframe annuity.
The bad news is out — this is no longer about the Q2 print. It's a credibility and guidance event. The setup is asymmetric in an unusual way: expectations are now on the floor after a historic crash, profitability and cash flow held up, and there are genuine bright spots (Red Hat, Confluent/HashiCorp, distributed infra, quantum).
The key questions to judge the call by: 1. How deep is the full-year cut, and is FCF reaffirmed? (FCF held = valuation floor holds.) 2. Is the miss framed as fixable timing/execution, or an admission of structural budget shift? 3. Is the mainframe/TP weakness a z17 timing air-pocket or the start of a cycle rolldown? 4. What does management say about capital returns, M&A, and the activist/break-up chatter?
A confident, specific guide with reaffirmed cash flow could mark a bottom; a hedged or absent guide would confirm the market's fear that this is more than one bad quarter.
(Note: figures above are IBM's preliminary/prior-guidance disclosures and market data through July 21, 2026; final reported results on July 22 could differ slightly, as IBM cautioned.)