Jim Cramer spent his July 6, 2026 Stop Trading segment pointing away from the obvious AI trade. While traders chase every hyperscaler capex beneficiary and GPU adjacency they can find, he told viewers to “check, without the cauldron of the data center, [they] should be looking at IBM.” His pitch leaned on a fresh Bank of America upgrade and a valuation that, in a market where AI names routinely trade north of 40x forward earnings, looks almost quaint.
IBM (NYSE:IBM | IBM Price Prediction) is the trade he wants you to make while the rest of the market is busy elsewhere.
The Cramer pitch, in his own words
Cramer’s setup was direct. “Bank of America raising price target, raising earnings per share. It’s got some of course AI. But it really is this great computer company,” he said, before landing on the number that matters. “And it sells at 22 times next year’s earnings. I think this one works.” He also acknowledged the elephant. “I know it got hit very badly when it reported, but I think it’s going to be a good, good idea.”
IBM printed a clean beat on April 22, then sold off anyway. Shares closed the filing day at $257.80, dropped roughly 10% within a week, then clawed back to $299.68 by Monday morning. BofA is now at $330 (raised from $315), citing software strength, Confluent synergies, and IBM’s dividend record. The forward P/E per Alpha Vantage is 23x, close enough to Cramer’s 22 to call it a match.
What’s actually inside the “boring” AI story
The AI part of IBM’s business is bigger than casual observers realize. The generative AI book of business had crossed $12.5 billion inception-to-date by year-end, with roughly four-fifths in Consulting and one-fifth in Software, and it has been accelerating from $7.5 billion in Q2 2025 and $9.5 billion in Q3 2025. That is real money attached to real workloads.
Q1 2026 gave the thesis teeth. Revenue of $15.917 billion, up 9.46% year over year, beat by 1.70%. Non-GAAP EPS of $1.91 versus $1.81 consensus made it the fourth consecutive EPS beat. Software grew 11.3% with Red Hat up 13% and Data up 19%. Infrastructure was the shocker. IBM Z mainframe revenue rose 51% year over year and segment margin expanding to 15.8% from 8.6%.
Arvind Krishna claimed on the call that IBM’s fully populated Z can now handle “about 450 billion inferences a day”, which is why banks are running fraud models directly on the transaction rail instead of shipping data out.
Then there is the ballast. IBM raised the dividend to $1.69 per share, the 31st consecutive annual increase, and the company has paid a quarterly dividend every year since 1916. That income floor does not exist in the data-center-darling universe. You can verify the Q1 numbers in the Q1 2026 8-K exhibit filed with the SEC.
Pressure-testing the 22x trade
Is 22x forward earnings actually cheap for what IBM does, or is it priced correctly for a company that grew Consulting only 4% in the quarter and carries elevated debt after Confluent? Free cash flow guidance calls for an approximately $1 billion year-over-year increase in 2026. Return on equity sits at 35.8%. Beta is 0.675, so you are getting AI exposure with less whip than the rest of the complex.
The bearish read has weight. Consulting is the largest slice of that $12.5 billion AI book, and consulting revenue growing 4% while the backlog is 30% GenAI raises a fair question about whether AI is expanding the pie or eating older services. Reddit conversation in June kept surfacing IBM in “forgotten tech stocks” threads, which is either the contrarian’s dream or the market telling you something.
Cramer’s call is coherent. A 2.25% dividend yield, a forward multiple in the low 20s, a real AI book compounding fast, and a mainframe cycle that will not quit. Whether that trade-off works depends on whether you are trying to win the next quarter or the next five years.
Contact [email protected] for any questions or corrections.
Top Gaining Stocks
Top Losing Stocks