Intel Just Lost A Third Of Its October Miracle, And The Bill Comes Due Thursday

Intel spent the first three weeks of October looking like the best trade on Wall Street. The stock spiked toward $123, riding a stack of headlines that read less like a semiconductor company and more like a midlife crisis with a happy ending: Elon Musk confirming Intel would still run the Terafab, Meta pledging day-zero support for Intel CPUs on its new Muse AI agent, a Tesla deal for Intel's next process node, analyst upgrades praising CEO Lip-Bu Tan. Then the mid-October Kiryat Gat explosion and a general AI-capex cold snap hit, and the rally bled out. Intel is now around $91.69 — down nearly a quarter from the September 23 peak, though only −2.4% over the last ninety days. The miracle didn't die. It just went quiet, and it's about to have to justify itself against a balance sheet that was never going to survive the hangover.

Let's be honest about what this stock actually is right now. It's not a turnaround story. It's a conviction bet on the idea that Intel can re-rate from a dying CPU vendor into an AI-infrastructure and foundry heavyweight — and that bet has had a spectacular run, up roughly 148% year-to-date — before the market started asking the question every bubble gets asked before it pops: at what price?

Here's the thing nobody in the bull narrative wants to say out loud. Intel's Q2 numbers were genuinely strong on the top line — revenue up about 25% year-over-year, its fastest growth in about fifteen years, data-center and AI revenue surging about 59% to $6.3 billion. But those were non-GAAP numbers, and the GAAP story underneath was a net loss with a per-share print of -$2.16 and a negative net margin. The company is growing revenue while losing money on the books. That's not a business. That's a pitch deck.

And the pitch deck is getting expensive. Intel Foundry is posting quarterly operating losses above $2 billion. Capital spending is on track to exceed $20 billion this year — more than the company operates in cash. That gap didn't fill itself. It got filled by a $20 billion equity raise in August, which is exactly the dilution the bulls quietly accept as the cost of the foundry dream, but which is doing real work right now to weigh on the share price. Price-to-sales near 8.5. You are paying up like a hyper-growth AI pure-play for a company that still isn't profitable.

So where does that leave us, as the next report looms? The calendar is itself a warning sign. The more sensible reading, given that Q4 wouldn't be reported until January 2027, is that this is a messy internal log. What matters is that Intel is reporting again within a week. Two earnings in seven days is not how a company with a clean narrative operates. It's how a company the market is forcing to prove twice over that the Q2 surge wasn't a fluke.

The consensus EPS hovering around $0.40 is not a high bar for a company that just printed negative $2.16. But "beating a low bar" doesn't move a stock that's priced for perfection and down 14.6% on the month. What moves it is guidance — specifically whether the CPU share bleed versus AMD and Arm slows, whether the foundry losses stabilize, and whether all those Terafab and Tesla and Meta headlines actually convert into revenue that shows up before the CapEx line.

My view? The bulls aren't wrong about the direction, but they're dangerously late to the party. The easy money in Intel was made between the Musk clarification and the mid-October selloff. What's left is the hard part: proving a company burning over $2 billion a quarter in foundry losses can fund its own transformation without diluting shareholders into oblivion. The AI narrative gave Intel a fifteen-year re-rating it never earned and may never earn. Right now the stock is a coin flip on Thursday's call, tilted slightly toward disappointment given how much good news already got priced in and worked its way through.

Buy the rumor, sell the news has never been a cleaner setup in this market. And Intel is about to deliver two pieces of news back to back. If management can point to margin stabilization and a foundry pipeline that isn't quietly subleasing its own fabs to TSMC, this thing stabilizes and maybe reclaims some lost ground. If they can't — if the guidance smells like more CapEx and more share dilution — the story could unravel fast, because everyone who bought the story already owns the exit.

The chart says the speculation is over. The fundamentals say the story isn't. The next week decides which one wins. I'd keep my hands off until Thursday's numbers tell me, in writing, whether Intel is a business or a brochure.