Brent Thill said Oracle's backlog rose 46% year over year, implying about $26 billion of net new contracts, and that the stock went into the print with everybody negative on it.
The bear case is the balance sheet and the OpenAI exposure. Thill's answer is that every one of the large cloud builders has now said the same thing about demand, that Oracle's own numbers were at the top of the range, and that the company's guidance is deliberately conservative.
"The RPO, which is their backlog, was up 46% year over year and implied $26 billion net new adds. They booked $30 billion more of AI contracts in the quarter. And everyone hates the stock."
Thill is Jefferies' Senior Software Analyst, with a buy rating and a $290 price target on the stock going into the quarter.
I listened to the full segment so you can skip it.
Here are the 3 calls that matter.
👤 Guest: Brent Thill, Senior Software Analyst at Jefferies
🎙️ Hosts: CNBC's Closing Bell Overtime anchors, Melissa Lee and Michael Santoli
📰 Published: 10 September 2026 on CNBC's Closing Bell Overtime
🟣 Apple Podcasts | 🔗 Episode page | ⏱️ 4 min
Key Takeaways
The backlog is up 46% year over year, implying about $26bn of net new contracts, with $30bn of AI contracts booked in the quarter
Oracle's cloud infrastructure grew 120%, ahead of what analysts were looking for
Thill's math is $20 of earnings power in three years at a 20 multiple, which is a $400 stock
The shares went into the print near his read of peak negativity, with a $290 target
He still ranks Microsoft, Amazon and Google ahead of it — Oracle is tier two, but one of only five companies he thinks will build this out
1. The Quarter Was Great
The desk put the obvious objection first: the stock is up on the print, but the guidance looked weaker than the quarter itself.
Thill read the guidance as deliberate caution rather than weakness. "Yeah. I mean, I think they're being conservative. The quarter was great."
The infrastructure number is the one he leads with. "If you look at OCI, which is Oracle cloud infrastructure, that was ahead of the bogey in 120% growth." OCI is the company's rented computing business, and the bogey is the figure analysts had set as the pass mark
The contracted revenue is the second. "The RPO, which is their backlog, was up 46% year over year and implied $26 billion net new adds. They booked $30 billion more of AI contracts in the quarter. And everyone hates the stock." RPO is remaining performance obligations — work sold but not yet delivered or recognized as revenue
He put the print in the context of the other cloud builders, saying that across the large infrastructure providers the message this cycle has been that business is booming or better, and that the argument about AI being too expensive to earn a return was thrown out the window
2. $20 Of EPS, A $400 Stock
Thill's valuation case is arithmetic rather than sentiment.
The earnings assumption comes from the company. "So we think again when you look at this company saying that they're going to get to $20 of earnings power in three years out. You can put a 20 multiple on it. It's a $400 stock."
He also noted the guidance actions: the earnings guide raised and the full-year figure of over $90 billion reiterated
What he thinks is in the price is fear rather than numbers. "And this fear of OpenAI not doing well and the debt load and all these other concerns are really weighing hard on the stock."
His summary of the quarter's figures was flat. Software still growing double digits, infrastructure at 120%, backlog up 46% — "These are really good numbers."
3. Tier Two, But One Of Five
The anchor pressed on the two structural differences between Oracle and the other hyperscalers: a less secure balance sheet, and a core business investors never liked as much.
Thill accepted the second point directly. Oracle was earning a 45% operating margin in a software business that was growing, but not at the rate of the others
His answer is scarcity of capable builders. "They're in tier two, but they're only going to be literally five companies that are going to build this out."
He does not call it the best way to own the theme. There are three other names he puts ahead of it — he named Microsoft, Amazon and Google as the firm's favorites — and the reason to own Oracle is execution against low expectations
His view of the setup was tactical. He said the stock had been pounded, everyone had turned negative, and it looked oversold going in
Thill's bottom line is that Oracle's contracted backlog and infrastructure growth put it among the five companies that will actually build the AI cloud, and that the stock was priced for the opposite.
Bonus Insights
Thill's price target and rating were stated on air — a buy with a $290 target — and his note into the quarter had argued there was peak negativity in the stock, which is the call the reaction tested
The anchor's framing of the multiple gap is worth keeping. Before this build-out phase, the core businesses of Amazon, Alphabet and Microsoft were more admired and more expensively valued than Oracle's, which is part of why the same demand story gets a different multiple
On the multiple itself, Thill argued the market is anchoring on the wrong number, saying that against what the company is on course to earn the multiple is not even that big
If this was worth your time, send it to someone who follows the name.
Get the latest market chatter as it happens: