Oracle’s stock has been on fire — but is it a sustainable surge? If not, this longtime rival could soon eclipse the database giant’s market value.
Database software titan Oracle (ORCL 6.67%) is soaring in 2025. As I write this on Sept. 18, the stock has gained 97% in six months. The company is knocking on the door of the trillion-dollar club, sporting a market cap of $855 billion. The company is landing billion-dollar cloud computing deals thanks to the ongoing artificial intelligence (AI) boom.
Oracle’s surge is impressive, of course. At the same time, it looks brittle. Can Oracle maintain its newfound identity as an AI hyperscaler? Investors certainly expect the good times to keep rolling for years, judging by the stock’s audacious valuation multiples. I think Oracle’s market cap could be smaller in five years — and still look expensive.
Longtime Oracle rival IBM (IBM 2.05%) is telling the opposite story. Big Blue is also building a large AI footprint, but Wall Street isn’t paying attention. IBM’s stock looks overlooked and undervalued with a $244 billion market cap.
Here’s why I think the average investor is missing IBM’s success story. By 2030, this stock could very well flip the script and grow larger than Oracle.
Big Blue is not flashy, just massive
You probably don’t see a ton of AI headlines from IBM these days — unless you work in enterprise-class data centers, of course. That’s not because Big Blue is sitting out the AI boom. The company is simply focusing on mega-client deals rather than press bluster.
This world-class innovator is building a full-stack, cash-generating AI play that most investors still haven’t priced into its stock. IBM is embedding business-class AI functions and features into everything, from big-box mainframes and data analytics to consulting services and hybrid cloud infrastructure.
It’s true that Oracle offers many of the same benefits at the moment. I’m not surprised at all. Larry Ellison’s company has a long history of emulating IBM’s best strategies. Back in 2012, when IBM refocused wholeheartedly on software, services, and cloud computing, Oracle was just trying to copy Big Blue’s old one-stop IT shop business model.
But it’s never quite the same. Oracle doesn’t build mainframe-compatible systems — it provides migration services to move applications away from mainframes instead. IBM owns the famous Red Hat Linux brand and the Red Hat Enterprise Linux (RHEL) platform. Oracle Linux is mostly a rebranded version of RHEL, optimized to run in Oracle’s cloud computing environment.
And it’s the same thing with business results. Oracle generates lower total sales than IBM, even in the current AI-driven surge. IBM collects robust and stable free cash flows, while Oracle’s cash profits are plunging into negative territory lately.
Image source: Getty Images.
IBM’s data quality moat
I haven’t even mentioned IBM’s quietly thriving AI business yet. There’s a lot to say; here are some highlights.
First, IBM starts from a different set of assumptions than most of its AI rivals. This company isn’t seeking quick wins or attention-grabbing headlines. Instead, it builds the AI systems a global enterprise would want to use, based on ironclad data integrity and audit-ready information flows. IBM’s watsonx won’t just give you the generative AI results you’re looking for, but will also show the original sources that produced this result.
Big Blue seeks a different type of customer, too. It’s all about the big names — Fortune 500 companies with large data warehouses and an absolute need for tight security. Sure, Oracle also loves these clients, but Big Blue’s engagement with this upper-crust demographic is on the next level.
That comes with some downsides, such as longer waiting periods before signing long-term contracts. IT directors in this category must put any new system through the paces with security and performance testing, not to mention explaining to upper management why IBM’s AI tools make sense. But when IBM lands a deal, the customer is usually locked in for many years, and unlikely to switch to another provider. That’s worth the delay in my book.
Furthermore, IBM’s deals don’t end with a software contract or a mainframe shipment. That’s just the start of a long partnership, usually involving plenty of IBM’s consulting services.
So IBM’s AI business is popping out of the financial starting blocks a couple of years behind the rest, but that’s OK. It’s a marathon, not a sprint. IBM’s generative AI contracts have soared from $2 billion in the summer of 2024 to $7.5 billion in the latest report, and the signings are accelerating.
Show me the cash: IBM vs. Oracle in a nutshell
You might expect analysts and investors to see value in IBM’s AI boom, but that process is also a bit slow. Today, IBM’s stock trades at 20 times free cash flow (while Oracle’s free cash flows are negative) and 22 times forward earnings (Oracle: 37).
So I won’t be surprised if IBM’s stock carries a larger market value than Oracle’s in 2030. Even if I’m wrong about that, Big Blue’s stock looks poised to perform over the next five years, but Oracle’s should slide back a bit. It already seems overdue for a price correction.
In short, IBM’s stock is a steal today but I wouldn’t touch Oracle’s overheated shares with a 10 gigabit network cable.