Hey there, bargain hunter. While the market fixates on Meta’s projected capital expenditure and Tesla’s margin slide, one company arriving at the same earnings window on Wednesday October 21 is making a very different argument about where AI money actually lands. IBM does not build GPU clusters. It invoices enterprises for the work of deploying them.
Scoreboard
IBM reports Q3 after the close Wednesday, October 21, 2026. Consensus sits at $2.86 per share on revenue of about $16.82 billion. That compares to $2.65 per share and $16.33 billion in the year-ago quarter, implying roughly 8% EPS growth and 3% revenue growth. IBM has beaten the consensus EPS estimate in seven of its last eight quarters.
The share price closed at $227.02 on October 12, roughly 32% below its 52-week high of $332.46. It bounced hard off that low: IBM closed at $234.12 on October 15. The stock is still well off its peak, but it has moved faster than the broader AI complex this month.
The Real Reason This Report Matters
The capex names, Meta and Lam Research most visibly, are being asked whether their enormous infrastructure spending will ever produce proportionate returns.
IBM is the control group. Its AI revenue does not depend on whether a data center gets built. It depends on whether a CFO signs a multi-year consulting contract. That is a different risk entirely.
How IBM Makes Its AI Money
IBM runs three main segments: Software (Red Hat, automation, data tools), Consulting, and Infrastructure (mainframes). The AI angle lives primarily in Consulting and Software. Consulting is where IBM is seeing early evidence that AI projects are moving beyond pilots, with generative AI representing about half of Consulting signings in the second quarter of 2026 and more than 30% of backlog.
IBM is not the only consulting giant where bookings and revenue have been telling different stories. Accenture’s fiscal fourth quarter showed how quickly enterprise AI consulting sentiment can reprice a stock down a third.
The AI work is landing in bookings well before it lands in the revenue line. That gap is the story. In Q2 2026, Consulting signings rose about 6% adjusted for currency. Consulting revenues increased 1% at constant currency, even as signings rose.
Data Section
- Q3 consensus EPS: $2.86, up ~8% year over year
- Q3 consensus revenue: about $16.82 billion, up ~3% year over year
- Software revenue estimate: $7.62 billion, implying year-over-year growth of 5.7%
- Consulting revenue estimate: $5.35 billion, implying year-over-year growth of 0.5%
- Consulting backlog at June 30: $30.8 billion
- Book-to-bill (trailing twelve months): approximately 1.05
- Free cash flow, first half 2026: $4.8 billion, flat year over year
- Full-year FCF target: free cash flow to increase by about $1 billion year over year
- Full-year revenue guide: 4% to 5% constant-currency growth
- Dividend yield: approximately 3.0%
Is It Cheap?
IBM’s trailing PE ratio is around 20x, with a forward PE near 19x. For context, peer multiples vary widely by source and timing, so treat any single-number comparison as a rough check rather than a verdict. IBM carries a premium to some large IT services peers, which is harder to justify if Consulting revenue fails to accelerate. But Wall Street’s average price target is $246.35, implying about 5% upside from recent levels, and the dividend creates a real floor beneath that math.
Bull / Base / Bear
Bull: Consulting revenue finally catches up to the backlog. GenAI signings compound through 2027, Software holds 6%-plus growth, and IBM raises its full-year FCF target. The stock rerates toward $260.
Base: IBM beats the $2.86 consensus modestly, holds guidance, and Consulting revenue edges above 1% constant-currency growth. The stock bounces 3%-5% on relief, then treads water while the backlog-to-revenue conversion plays out over multiple quarters.
Bear: Consulting revenue fails to reach the low to mid-single-digit growth management guided, while signings keep climbing, meaning the backlog is not converting. That is the tell that enterprise AI spend is stalling rather than shifting. The stock retests $200.
Action Plan
Do not buy ahead of the report hoping for a beat. Options traders are pricing roughly an 8% move in either direction. The better trade: wait for the Consulting revenue number. If it clears 2% constant-currency growth and management holds the FCF guide, add a starter position with a 48-hour lag. Scale in a second tranche if the stock pulls back on a broader market selloff driven by Meta or Tesla reaction, not IBM’s own results.
Cheap Investor Checklist
- Consulting revenue growth at or above 2% constant currency in Q3
- GenAI percentage of Consulting backlog holds above 30%
- Full-year FCF guide maintained at approximately $1 billion year-over-year increase
- Book-to-bill stays above 1.0 for trailing twelve months
- Software segment sustains above 5% growth (Red Hat traction is the driver)
- Management updates or replaces the retired inception-to-date generative AI revenue total
- Operating pre-tax margin expansion on track for 100 basis points full year
- No further full-year revenue guide reduction below 4% constant currency
Bottom Line
If IBM’s Consulting revenue starts closing the gap with its own backlog, the stock is meaningfully cheap at around $232 with a 3% dividend cushion. If the signings keep compounding while revenue sits at 1%, you are holding a well-managed company with a conversion problem, and patience will be tested for another quarter. Wednesday night answers the question the backlog has been raising since June.
