Intel posted its fastest revenue growth in roughly 15 years on Thursday evening, July 23, doubled the earnings estimate CNBC cited, guided above expectations for Q3, and still lost 7.89% in Friday's session. Intel is a semiconductor company that designs and manufactures its own chips, and its foundry arm, the business that builds chips for outside customers, is the part of the story the market refused to pay for. The reaction was not a slow fade either. It was a full round trip from after-hours euphoria to a close below the pre-earnings price.
INTC snapshot (Friday, July 24 close):
- Close: $92.32, down 7.89% on the session
- Q2 revenue: $16.128 billion, up 25% year over year, the fastest growth in about 15 years per CNBC
- Adjusted EPS: $0.42, double the $0.21 consensus CNBC cited
- Data Center and AI segment: up 59% year over year, per TradingKey
- Volume: 179.1 million shares, roughly 36% above average, per Motley Fool
A quarter this strong getting sold this hard needs a better explanation than the headline numbers, and it has one.
The Numbers Intel Printed Against the Estimates
Start with what actually hit the tape, because the beat was not marginal on any line. Revenue came in about $1.7 billion above the consensus figure CNBC cited, and the earnings beat was large enough that "doubled" is literal.
Metric | Consensus (per CNBC) | Actual | Verdict |
Q2 revenue | $14.42 billion | $16.128 billion, +25% YoY | Beat by roughly $1.7 billion |
Q2 adjusted EPS | $0.21 | $0.42 | Doubled the estimate |
Q3 revenue guidance | $15.1 billion expected | $15.8-16.8 billion, midpoint $16.3 billion | Guided above |
Q3 adjusted EPS guidance | $0.27 expected | $0.38 | Guided above |
One caveat belongs on the record before anyone repeats the "doubled estimates" line. Consensus trackers did not agree with each other heading into this print, and figures of $0.10 and $0.22 also circulated across platforms. The doubling claim holds against the $0.21 number CNBC cited, which is why every row in that table carries the attribution. If you want to check the underlying filings yourself, Intel's 8-K filings on SEC EDGAR carry the guidance ranges as reported.
The growth engine was exactly where you would want it. The Data Center and AI segment grew 59% year over year per TradingKey, which is the segment that determines Intel's relevance in the AI buildout. A company that spent a decade losing server share printed a data center number that looks like a share-gainer's. And the Q3 guide, above expectations on both revenue and earnings per share, says management expects the momentum to continue rather than treating Q2 as a one-off.
How a 12% After-Hours Pop Became an 8% Friday Loss
The sequencing is the story here, so keep the timeline straight. Intel reported after Thursday's close, and the stock jumped about 12% in the after-hours session, briefly trading near $112.70. That print exists only as a Thursday-evening snapshot. By the time Friday's regular session ended, INTC sat at $92.32, down 7.89% from the $100.23 pre-earnings close. A stock that was up double digits on the numbers finished the next day below where it traded before anyone saw them.
That shape of reversal carries information. An after-hours pop that fades at the open can be dismissed as thin-liquidity noise, but this one kept selling all session and closed near the lows on 179.1 million shares, roughly 36% above average volume per Motley Fool. Heavy volume on a full retrace of an earnings pop means large holders used the liquidity of the good news to exit. Retail bought the beat on Thursday night. Someone much bigger sold it to them all day Friday.
The pattern is common enough that experienced earnings traders treat the after-hours print as an opening bid, not a verdict. The market needed one full session to read past the headline beat, and what it found underneath changed the trade entirely.
The Foundry Question the Beat Could Not Answer
Coverage of Friday's session pointed to two drivers behind the reversal, and neither disputes the quarter. The first is foundry customers. Intel's turnaround case rests on 18A, its most advanced manufacturing process, winning committed external volume, meaning outside companies paying Intel to fabricate their chips. The Q2 print showed Intel's own products selling well. It did not resolve the question of who, beyond Intel itself, is filling those fabs at scale, and investors who came for foundry proof left without it.
We have been tracking this exact fault line for months. The stock's last major breakdown came on an 18A timeline delay, which we covered in our breakdown of the Intel 18A crash, and our July 20 Intel earnings preview argued before the print that the 18A story had reversed and the earnings reaction would hinge on foundry evidence rather than the headline numbers. Friday validated that read in the least pleasant way possible for longs.
The second driver is AI spending durability. The 59% data center growth is only worth a premium multiple if the capex wave funding it keeps rolling, and this week gave investors their first real reason to interrogate that assumption. When the market starts doubting the spend itself, a blowout quarter reads as a peak. That is the difference between the same number rallying a stock in June and sinking it in July.
The Rest of the Chip Complex Traded Lower With It
Intel did not fall in isolation, which matters for how you interpret the move. AMD dropped 3.29% to $521.95 on Friday, Micron fell 6.99% to $920.95, and NVIDIA slipped 0.92% to $206.84. The Nasdaq lost 0.64% to 24,975.82 and booked its second straight weekly loss, down 2.1% for the week. You can track the sector's aggregate move on the PHLX Semiconductor Index page on Nasdaq, and the picture it paints is a sector-wide repricing that swept up strong and weak reports alike.
The context is Thursday, July 23, which was the week's real damage session. Tesla fell 14.52% that day and Alphabet's capital-spending shock rattled the entire AI trade, so Intel released the best quarter it has printed in years into a market that had started questioning who pays for the AI buildout. Micron's 6.99% Friday drop is the tell. A memory supplier with no Intel-specific news falling nearly as hard as Intel says the market was selling AI capex exposure broadly, a dynamic we flagged as the central risk in our Micron price prediction through 2030.
NVIDIA's modest 0.92% slip against Intel's 7.89% drop is its own signal. The market is not abandoning AI semiconductors. It is sorting them by how much proof each one has attached to its story, a sorting process we walked through in our NVIDIA stock outlook. NVIDIA sells the shovels everyone is verifiably buying. Intel is still asking the market to underwrite a foundry future it has not fully demonstrated, and weeks like this one show what that gap costs.
What INTC Traders Should Watch From Here
The trade now has clean reference points, which is more than most post-earnings situations offer. The $100.23 pre-earnings close is the line that separates interpretation from noise. A reclaim of that level would say the market has re-read the quarter and decided the foundry worry was overdone. Continued closes below $92.32, especially on elevated volume, would say the distribution that started Friday is still running.
Catalysts are stacked close together. With the Fed's July 29 decision looming, rate-sensitive growth names like Intel carry macro risk into midweek, a setup we cover fully in our Fed decision preview. Beyond that, the next scheduled read on the foundry question is management commentary, and any announcement naming a committed external 18A customer would reset the entire debate in a single headline. That is the asymmetry in both directions. The stock now trades below its pre-earnings level with a doubled estimate behind it, so the bad news is at least partially priced, while the specific proof the market wants remains a binary future event.
Position sizing matters more than usual in that setup. A stock that can round-trip 12% of after-hours upside in one session can move violently on the next foundry headline, in either direction.
Frequently Asked Questions
Why did Intel stock drop after good earnings?
Intel stock fell 7.89% on Friday, July 24, 2026 because investors focused on unanswered questions about external foundry customers and the durability of AI capital spending rather than the Q2 beat. The report proved Intel's own products are selling, but the turnaround thesis requires outside companies committing to Intel's 18A manufacturing at scale, and the quarter did not demonstrate that.
Did Intel beat earnings expectations in Q2 2026?
Intel beat on every headline line in Q2 2026, reporting revenue of $16.128 billion against the $14.42 billion consensus CNBC cited and adjusted EPS of $0.42 versus the $0.21 expected. Revenue grew 25% year over year, which CNBC noted was Intel's fastest growth in about 15 years, and the Data Center and AI segment grew 59% per TradingKey.
What is Intel's guidance for Q3 2026?
Intel guided Q3 2026 revenue to a range of $15.8-16.8 billion with a midpoint of $16.3 billion, above the $15.1 billion analysts expected, per CNBC and the company's 8-K filing. Adjusted EPS guidance of $0.38 also came in ahead of the $0.27 estimate, so the above-consensus quarter was paired with an above-consensus outlook.
Is Intel stock a buy after the drop?
Intel stock at $92.32 trades below its pre-earnings price despite a quarter that doubled the CNBC-cited earnings estimate, which makes the pullback attractive on the reported numbers alone. The honest answer depends on foundry progress, because without named external 18A customers the market has shown it will discount even excellent results, so the position is a bet that proof arrives before patience runs out.
Bottom Line
Intel printed the quarter the bulls asked for and the stock closed 7.89% lower anyway, which tells you the market has moved past the income statement and onto the foundry roadmap. If INTC reclaims and holds $100.23, treat the selloff as positioning that overshot and let the above-consensus Q3 guide do the work. If the stock keeps closing below $92.32 on heavy volume, the Friday sellers are still in control and there is no rule that says distribution stops after one session. If Intel names a committed external 18A customer, the debate resets overnight and the doubled estimate becomes the floor of the story. Until one of those three things happens, this is a headline-driven tape where the next foundry announcement outweighs the last earnings report.
This article is for informational purposes only and does not constitute financial or investment advice. Cryptocurrency trading involves substantial risk. Always conduct your own research before making trading decisions.
