Intel (INTC) Q3 2026 Earnings: How $140B Vanished in a Day

How does a household-name blue chip worth almost $490 billion lose about $140 billion of market value between one closing bell and the next? That is the question Intel left behind on 30 October 2026. The shares closed at $65.76, down 28.58% or $26.31 from $92.07, per Motley Fool data. In a single regular session, the market removed roughly six times the $23 billion Intel raised in its public equity offering this year.

The astonishing part is what came first. Intel reported Q3 2026 results on 29 October 2026 after the US close, and they were above its own guidance. Revenue of $16.5 billion rose 21% and was in line with LSEG consensus of $16.5 billion. Non-GAAP EPS of $0.52 beat the $0.40 LSEG consensus and the company’s $0.38 guide. Data center revenue rose 78%. After the numbers came out, the shares fell more than 20% in extended trading, per CNBC, and the next day was worse.

CNBC called it Intel’s worst day since the 1970s. In 2024 CNBC reported that the 26% fall of August 2024 was the second-worst day ever, behind a 31% drop in July 1974, so this session now takes second place on that yardstick. Yahoo Finance daily data back to 1980 shows no worse day. Volume reached about 450 million shares, more than four times the three-month average of about 108 million.

The quarter did not break the stock. The next quarter did.

1. Intel Q3 2026 Earnings Scorecard

Metric	Actual	Consensus	Company guidance	YoY	Result
Revenue	$16.5B	$16.5B	$15.8B to $16.8B	+21%	In line, above guide midpoint
EPS (GAAP)	$0.60	n/a	$0.31	Down from $0.90	Above guide
EPS (non-GAAP)	$0.52	$0.40	$0.38	Up from $0.23	Beat
Gross margin (non-GAAP)	44.1%	n/a	42.0%	Up from 40.0%	Above guide
Operating margin (non-GAAP)	19.6%	n/a	n/a	Up from 11.2%	Expanded
Adjusted free cash flow	$1.6B	n/a	n/a	Up from $0.9B	Improved
DCAI revenue	$7.3B	n/a	n/a	+78%	Standout
Q4 revenue guidance	$15.0B to $16.0B	$17.1B	New guide	Midpoint +13%	Miss
Q4 EPS guidance (non-GAAP)	$0.25	$0.43	New guide	Up from $0.15	Miss
Q4 gross margin guidance (non-GAAP)	38.0%	n/a	New guide	Up from 37.9%	Down 6.1 points on Q3

Source: Intel earnings release, 8-K dated 29 October 2026, and prior-quarter releases; consensus per LSEG via CNBC, as of 29 October 2026. “In line” means within 1% of consensus. Data as of 30 October 2026. Verify before acting.

GAAP EPS is flattered by a $1.0 billion mark-to-market gain on the escrowed shares tied to the US government’s stake, worth about $0.17 a share. Non-GAAP EPS excludes that gain, which is why it sits below the GAAP figure this quarter.

2. The Beat That Did Not Matter: Revenue and Segments

Read on its own, this was a very good quarter. CFO David Zinsner described “our strongest data center revenue growth in 15 years,” and the company said it exceeded guidance on revenue, gross margin and EPS. Then the same report handed investors the reason to sell.

Segment	Revenue	YoY	Share of total
Client Computing and Physical AI (CCPG)	$8.1B	-5%	49%
Data Center and AI (DCAI)	$7.3B	+78%	44%
Intel Foundry	$6.7B	+58%	Mostly intersegment
All Other	$0.7B	-30%	4%

Intel Foundry revenue is mostly sales to Intel’s own product groups, so it is eliminated on consolidation. External foundry revenue was $411 million, up from $32 million a year earlier, per the 10-Q.

DCAI operating income of $3.0 billion overtook CCPG’s $2.3 billion, while Intel Foundry lost $1.9 billion. CCPG, the PC franchise, is still the largest revenue segment, but it is shrinking. That detail turned out to matter more than everything else in the table.

3. Margins, Profit and Cash Flow

Non-GAAP gross margin has climbed for three straight quarters: 37.9% in Q4 2025, 41.0% in Q1 2026, 41.8% in Q2 and 44.1% in Q3. Non-GAAP operating margin reached 19.6%, against 17.2% in Q2. On a margin chart, Q3 looks like a turnaround that was finally working.

Cash is the thinner part of the story:

• Operating cash flow was $5.1 billion and adjusted free cash flow was $1.6 billion.
• Cash and short-term investments total about $54 billion, against total debt of about $50 billion.
• The $23 billion equity offering helped lift diluted weighted shares to 5,307 million from 4,531 million a year earlier, about 17% more.
• Gross capital spending is on track for about $20 billion in 2026, and Zinsner expects 2027 to be “up strongly.”

That last line collided with a blunt answer on the call. Asked about 2027, Zinsner said: “Probably not a dynamic where we are going to have positive free cash flow next year.” Investors holding a stock that had already multiplied in value heard that capex is rising while near-term earnings fall.

4. The Guide That Broke the Stock

This is where a beat became a historic collapse. For Q4 2026, Intel guided:

• Revenue of $15.0 billion to $16.0 billion, against LSEG consensus of $17.1 billion. The $15.5 billion midpoint is about $1.6 billion short, and even the top of the range misses by $1.1 billion.
• Non-GAAP EPS of $0.25, against consensus of $0.43, and less than half of Q3’s $0.52.
• Non-GAAP gross margin of 38.0%, against 44.1% in Q3. GAAP gross margin of 37.1% and GAAP EPS of $0.10.

Think about the speed of that. In one quarter, 6.1 points of gross margin disappear and non-GAAP earnings are cut by more than half. Zinsner said the decline “primarily reflects lower overall revenue against our fixed cost base.” That is the cruel arithmetic of owning factories: when volume drops, the costs stay.

Fab 28 explains part of it, and it should be reported soberly. CEO Lip-Bu Tan said “an unfortunate and tragic accident led to an extended shutdown in our Israel factory.” Zinsner put the Q4 impact at “roughly $300 million to revenue and 200 basis points to gross margin,” almost all in 10-nanometer server products, with a similar impact likely in Q1 2027. He added that the margin effect is “really more of a Q4 event.” So about one third of the gross margin fall is linked to that shutdown, and about two thirds is not.

Year on year, the guide is not disastrous: the midpoint is 13% above Q4 2025 revenue of $13.7 billion, and EPS of $0.25 compares with $0.15. The stock was not priced for that comparison. It was priced for acceleration.

5. The AI Irony: Memory Squeezes PCs, Supply Caps Servers

Here is the insight that makes this story so striking. The same AI boom lifting Intel’s data center business is hurting its PC business. AI demand is pulling on memory supply, and Tan said “our customers began signaling a PC slowdown based on the memory prices and availability.” Our Micron HBM analysis explains why memory has become such a bottleneck.

PCs are falling fast

The numbers from the call are severe:

• Zinsner expects the 2026 PC unit market to fall “approximately 12%, with second-half units down more than 20%.”
• For 2027, Intel is “planning for the PC unit TAM to decline roughly 10%.”
• CCPG revenue is expected to “decline greater than 20% sequentially” in Q4.
• Zinsner, answering Bernstein’s Stacy Rasgon: “we are not optimistic about next year in terms of memory.”

That is a second straight year of falling PC units, for the segment that still brings in almost half of revenue. Benchmark’s Cody Acree summed it up on CNBC: Intel is “still PC dependent.”

Sold out of what is working

Data center should fill the hole, and that is the frustration. Zinsner said “demand for DCAI products continues to meaningfully exceed available supply.” He expects DCAI revenue to grow by a mid-teens percentage in Q4. On our arithmetic, that takes DCAI to about $8.4 billion while CCPG falls below about $6.5 billion, so data center would become Intel’s largest segment. The swing is real, but it is not big enough to offset the PC decline.

The limits are physical: Intel 3 wafers, and what Tan called “an industry-wide shortage of advanced organic substrates for our EMIB packaging,” EMIB being Intel’s chip packaging technology. Zinsner said “the first half of 2027 is likely to be tighter in terms of wafers and substrates.” Intel is adding 40% more substrate capacity, but “a lot of that does not hit us until the back half of the year.” He expects the industry to stay “supply constrained throughout 2027 and into 2028.”

So investors are left with a long-term promise and a near-term gap. Tan sees the server CPU market rising “from $40 billion to more than $200 billion over five years,” yet Intel cannot ship enough today to benefit fully. Demand is not the problem. Delivery is.

6. What Changed in the Investment Thesis

So how can this happen to a company of Intel’s size in one day? Intel lost its Dow seat to Nvidia in November 2024, but it remains an S&P 500 and Nasdaq-100 giant, which makes the scale harder to believe. Numbers explain it better than adjectives. Intel closed 2025 at $36.90 and reached a high of $142.35 around the end of June 2026, a gain of about 286% in six months. It had already slid 35% from that high to $92.07 before this report. A stock priced for a comeback has a long way to fall when the comeback story cracks. The speed still stuns: about 8.5% of all Intel shares changed hands in that one session.

Confirmed. CPUs are back in demand. DCAI grew 78%, and supply, not customers, is the limit.

Weakened. The idea that Intel had outgrown the PC cycle. It has not, and CCPG’s expected Q4 fall shows how quickly that revenue can move.

Still unproven. Intel Foundry. It lost $1.9 billion in Q3, and Tan named no 14A customers, saying only that he is “highly encouraged” by progress with “multiple potential customers.” He said a couple of 18A-P customers should be ready to announce within two quarters. The 14A design kit reached version 0.9 in October, and Tan expects the next step by Q1 2027.

The damage spread beyond Intel on the same day. AMD fell 10.88% and Nvidia 4.22%, per Motley Fool, and the Nasdaq Composite slid 2.9%, per CNBC. The US government’s roughly 10% stake lost about $12 billion, and Nvidia’s stake lost about $6 billion, per CNBC. The week had already shown how unforgiving this market is, as our Alphabet Q3 2026 analysis and Oracle Investor Day review both show from the cash flow side. Our Marvell data center analysis offers wider context on AI silicon.

Our view: the fall reflects both a real cut to near-term earnings and a crowded position heading for the exit at once. The size of the move looks driven by what was priced in, more than by any collapse in the long-term demand story. That is a judgement, not a fact, and the next two quarters will test it.

7. Valuation After the Fall

Even after losing about $140 billion, Intel is not obviously cheap. The working on that headline number: $26.31 multiplied by 5,315 million shares outstanding (10-Q cover, 23 October 2026) is about $139.8 billion, taking implied market value from about $489 billion to about $350 billion. That is our own calculation.

• Forward P/E: about 40x after the drop, against 53x at the 29 October close, per TIKR data cited by Yahoo Finance.
• Trailing non-GAAP P/E: about 48x on $1.38 of EPS over the last four quarters, our arithmetic.
• Annualised Q4 guide: about 66x on $1.00, a low quarter annualised, so a stress test rather than a fair measure.
• Price to sales: about 5.3x annualised Q3 revenue.

On Wall Street, 17 analysts cut their targets, per Finbold. Bank of America’s Vivek Arya moved to Neutral from Buy and cut the target to $100 from $145, reducing his 2027 EPS estimate by 16%, per TipRanks. These are opinions, dated 30 October 2026, not forecasts we endorse.

8. Bull Case and Bear Case After the Results

🐂 Bull Case

• DCAI revenue rose 78%, and demand still exceeds supply into 2028.
• Management sees the server CPU market rising from $40 billion to more than $200 billion over five years.
• The PC decline is tied to memory prices and availability, which have historically been cyclical.
• Fab 28’s margin impact is expected to be mainly a Q4 event.
• Liquidity of about $54 billion, and backing from the US government and Nvidia, which bought 215 million shares at $23.28.

🐻 Bear Case

• Non-GAAP EPS is guided from $0.52 to $0.25 in one quarter.
• Management is planning for PC units to fall about 10% again in 2027, with no sustained recovery until memory improves.
• Supply tightens further in the first half of 2027, so DCAI cannot quickly fill the gap.
• Capex rises in 2027 and positive free cash flow is “probably not” on the cards.
• Intel Foundry lost $1.9 billion and has no named 14A customer, while BofA says Intel is losing ground to AMD and Arm-based rivals.
• At about 40x forward earnings and still up more than 70% in 2026, the stock leaves room for further disappointment.

9. Final Thoughts: Falling Knife or Reset?

It is hard to overstate how unusual this was. Intel’s Q3 2026 earnings beat expectations and its shares still suffered their worst day in about half a century. Yet perspective matters: even at $65.76 the stock is up about 78% from its 2025 close, while sitting about 54% below its $142.35 high.

For anyone tempted to catch the knife, history offers no comfort either way. Intel has had 23 one-day falls of 10% or more since 2000. Holding for 60 trading days from the close was positive only 11 times, or 47.8%, a figure from TradingKey that we reproduced from Yahoo Finance price data. The median return was about minus 1%, with a range from minus 31% to plus 47%. This is history, not a forecast.

The biggest positive is real AI CPU demand. The biggest concern is that Intel cannot yet turn it into enough supply while its PC business shrinks. At the Q4 report, expected in early 2027, watch whether gross margin rises from 38%, substrate and Intel 3 supply, named 18A-P or 14A customers, memory pricing, and the 2027 capex figure. For a look at the foundry landscape, read our GlobalFoundries analysis, and subscribe to the OneMoreStock YouTube channel at www.youtube.com/@onemorestock.

FAQ

Did Intel beat earnings expectations in Q3 2026?

Yes on EPS, and in line on revenue. Non-GAAP EPS of $0.52 beat LSEG consensus of $0.40 and Intel’s $0.38 guide. Revenue of $16.5 billion matched LSEG consensus and landed above the midpoint of Intel’s $15.8 billion to $16.8 billion guide.

Why did Intel stock crash after earnings?

The forward numbers, not the quarter. Intel guided Q4 revenue of $15.0 billion to $16.0 billion against consensus of $17.1 billion, and non-GAAP EPS of $0.25 against $0.43. It also expects PC units to fall more than 20% in the second half because of memory prices and availability, and it cannot supply enough data center chips to fill the gap.

How much value did Intel lose in one day?

About $140 billion. The shares fell $26.31 on 30 October 2026, and with 5,315 million shares outstanding that is about $139.8 billion by our calculation, taking implied market value from about $489 billion to about $350 billion.

Was this Intel’s worst day ever?

Not quite. CNBC called it the worst day since the 1970s. CNBC reported in 2024 that the worst day on record was a 31% drop in July 1974, so the 28.58% fall on 30 October 2026 ranks second on that basis, ahead of the 26% fall of August 2024.

What is Intel’s guidance for Q4 2026?

Revenue of $15.0 billion to $16.0 billion, non-GAAP gross margin of 38.0%, GAAP gross margin of 37.1%, non-GAAP EPS of $0.25 and GAAP EPS of $0.10. The gross margin and EPS figures are based on the revenue midpoint of $15.5 billion.

Is Intel stock a buy after the crash?

That depends on your horizon and your tolerance for a difficult 2027, and it is not a question one trading day can settle. The AI CPU demand looks real, but earnings are falling, supply tightens in the first half of 2027, and management does not expect positive free cash flow next year. Long-term investors should watch supply, gross margin and foundry customers rather than the size of the drop.

Financial Disclaimer:

This article is for educational and informational purposes only and should not be considered financial or investment advice. Investing involves risk, including the possible loss of capital. Always conduct your own research and consult a qualified financial adviser before making investment decisions.