Oracle burned $5.4 billion of free cash flow in the quarter, which the Fast Money desk pointed out was twice as good as the $10 billion of burn the street had penciled in.
The stock went into the print with nobody defending it. What the desk kept returning to was not the quarter but the balance sheet behind it: $167 billion of debt against contracts the traders are not sure the company can execute.
"This is a company that has $167 billion in debt."
The people saying it trade for a living: the Fast Money desk sat with the numbers within minutes of the release, priced the options market around it, and named their own positions at the end of the hour.
I listened to the full segment so you can skip it.
Here are the 5 calls that matter.
🎙️ Host: Melissa Lee, who anchors CNBC's Fast Money
👥 Also on: the Fast Money traders, of whom Tim Seymour, Karen Finerman and Dan Nathan are named on air in the final-trade segment; Alex Sherman, CNBC's sports-business reporter, reporting from Melbourne; Seema Mody, CNBC reporter, on the Oracle release
📰 Published: 10 September 2026 on CNBC's Fast Money
🟣 Apple Podcasts | 🔗 CNBC | ⏱️ 12 min
Key Takeaways
Oracle's free cash flow was negative $5.4B against a street expecting about negative $10B
The desk read the print as good enough to lift the whole software trade, not just the stock
$167B of debt and rising remaining performance obligations are the desk's actual worry
There is no guarantee, one trader said, that the company can fulfill what it has signed
Options were pricing a 12% move in either direction, which made calls too expensive to buy
Apple is the defensive trade in a weak tech tape, even at a higher multiple
One trader is long and not selling, but will wait for version two or three of the foldable
CNBC's own valuations put the average NFL franchise at $10.4B, up 5% in a year
Only 10% of a team may be owned by a private-equity firm, and the commissioner says the league is nowhere near that limit
1. Oracle's Free Cash Flow
The setup was as one-sided as the desk has seen. Melissa Lee: "And maybe the setup for the stock was perfect going into this report in that everybody was negative." One trader noted Steve Grasso had defended it on the show the night before.
The cash-flow number was bad in absolute terms and good against expectations. "You know the free cash flow which has been a disaster wasn't good at 5.4 billion negative. But it was twice as good as the street was expecting, if that makes sense because the street was looking for -$10 billion." Margins were slightly better as well.
The read-through the desk cared about was the sector, not the stock. "And I will continue to say I think the software trade can go higher."
Capital raising is off the table for now, which the desk noted and then questioned. Lee pointed out the company had completed its offering, and a trader said the important line was that there are no new plans to raise capital now — before adding that he wonders how long "for now" lasts.
Capital spending is the number still missing. "I believe the expected capex for this fiscal year is $90 billion or so." The desk said they want to see where the company lands, having watched everyone else surprise to the upside.
The options market made the trade expensive either way. A trader said the setup in options was "pretty crazy," with a "12% implied move in either direction," and that he was glad not to have paid for calls because the upside was very expensive: "And in fact, I don't think you're getting rewarded here."
2. $167B of Debt
The balance sheet is the desk's actual position on Oracle. "This is a company that has $167 billion in debt." The credit market has noticed: one trader said the debt is an integral part of the story and that the credit default swaps reflect higher concern about it, though "But at the moment this is positive."
Signed contracts are not delivered contracts. On the growing remaining performance obligations: "There is no guarantee that they are going to be able to fulfill those obligations." The desk's doubt was about execution history rather than demand — "I don't think we ever questioned the demand. The question is, can they fulfill it?" — with cloud infrastructure revenue up 121% as the evidence of demand.
The chain the desk is watching runs through the borrowers. A trader said that if OpenAI falters on its side of the demand, what follows starts with the heavily indebted companies rather than with the chipmakers.
Lee brought two analysts' view to the table against the desk's. "I spoke to two analysts on closing bell overtime Gil Luria of D.A. Davidson and Brent Thill of Jefferies, and they're saying this is the one company that is building out and is not getting full credit for that backlog." Her comparison, from them, was that CoreWeave, which has no profitability, and Microsoft both get full credit for theirs.
The desk's partial concession on the debt. A trader said that when Oracle proceeds with projects further out, it sometimes gets paid upfront or coincident with the spending, "so that helps a little bit," and that he thinks of Oracle as a bellwether for the group.
3. Apple's Defensive Bounce
Lee's framing of the day: the phone folds, the stock did not. "The new iphone may fold, but the stock sure didn't. At least not today." Apple rose over 3.5%, leading the Magnificent Seven, with early reviews of the foldable largely positive.
Two sell-side reactions the show read out. "Morgan Stanley analyst Erik Woodring noting that even with a $2,000 starting price, the duo is still cheaper than some feared." Melius pointed at the thinner profile and said "this foldable on Ozempic is going to be a hit."
The desk's own buying plan is to wait a generation or two. "As our chief GLP correspondent here, I can tell you that I'm not buying version one, but I'll buy version 2 or 3."
The bounce looked defensive rather than enthusiastic. "Part of today's bounce, I felt a little defensive though. I mean, in the sea of like tech weakness, right? Apple is sort of the ultimate defensive play." The same trader called it absolutely a stock that rallies on a day like this one.
Long, and honest about the valuation. "So I'm long Apple and I'm not selling Apple. And I understand the multiple gets a little bit more egregious." He added that the company has yet to play its real card, and that what has to change is Siri and the whole relationship between the phone and AI.
4. NFL Valuations at $10.4B
Alex Sherman reported from the first NFL game ever played in Australia. "I come to you from the future. Yes, it is Friday morning here in Melbourne." He was on the roof of a hotel by the Melbourne Cricket Ground: "It seats 100,000 people where tonight in the U.S. the San Francisco 49ers play the Los Angeles Rams in the first ever game in Australia."
The international schedule is the league's growth plan. "This is part of the NFL's big expansion plan. Nine international games this year. Commissioner Roger Goodell has said he wants 16."
CNBC's own valuation work is the investable part. "CNBC is out with its valuations. The average NFL team rose 5% from just a year ago." And: "The average NFL franchise, valued at $10.4 billion."
The private-equity cap is not binding, on the league's own account. Sherman said the current rule is that "Only 10% of a team can be owned by a pe firm," and that he asked Goodell whether prices have risen enough to force the cap higher. Goodell's answer, in the recorded interview Sherman played, was that the league is not close to the capacity of those caps either individually or collectively, that the decades-old worry about too few buyers for franchises is not what he sees, and that demand to invest in sports and in the NFL in particular is a reflection of the league's strength.
The distribution is the other half of the story. The game aired globally on Netflix, and Sherman said "It's one of five NFL games that will air on Netflix this season."
5. The Final Trades
Tim Seymour bought the NFL through the streamer. "If you're not aware, the first three letters of Netflix are NFL." His final trade was Netflix.
Karen Finerman went outside the AI complex. "So if you want to diversify away from the AI trade somewhat, I really like letter c, citibank." Her final trade was Citigroup.
Dan Nathan agreed with the premise. "First of all, I feel the same way actually about diversifying away from the AI trade." The desk's other final trades that day, per the segment's own title, were Zscaler and Oracle.
Bonus Insights
The desk's precedent for a product-driven Apple rally is June 2024. A trader recalled Gene Munster on the show the previous night saying the foldable would be big for the stock in the near term — "I want to give Gene Munster a shout out because he was on last night with us" — and compared it to the launch of Apple Intelligence: "Stock went up 20% in a straight line over the next month." The point he drew from it is that investors buy the inflection point before it is modelable in next year's numbers.
One trader's phone was dying on air as he made the case, which he offered as his own reason to buy the foldable.
The desk's bottom line is that Oracle's quarter was good enough to keep the software trade going but not good enough to settle the question the balance sheet poses — $167 billion of debt against obligations the traders doubt the company can deliver — while Apple, at a richer multiple, is where the same desk is hiding from a weak tech tape.
Products, Companies & Tools Mentioned
Oracle (Negative $5.4B free cash flow against a street looking for about negative $10B, cloud infrastructure revenue up 121%, and $167B of debt)
Apple (Up over 3.5% the day after the foldable reveal, which the desk read as a defensive bid in weak tech)
Netflix (Streamed the first NFL game in Australia and has five NFL games this season; Tim Seymour's final trade)
NFL (Nine international games this year against a stated goal of 16, an average franchise valued at $10.4B by CNBC, and a 10% cap on private-equity ownership)
Citigroup (Karen Finerman's final trade, as a way to diversify away from the AI trade)
CoreWeave and Microsoft (The comparison two analysts made to Oracle: both get full credit for their backlogs, Oracle does not)
OpenAI (The demand behind Oracle's contracts, and the single point of failure the desk named)
If this was worth your time, send it to someone closer to the industry than you are.
Get the latest market chatter as it happens: