Intel Reports Q2 Revenue Up 25% and Beats Profit Forecasts, but Posts $11 Billion GAAP Loss Tied to Government-Stake Charge
Intel's adjusted profit, revenue and guidance topped Wall Street estimates on data-center and AI demand, while a non-cash accounting charge linked to the U.S. government's equity stake drove a large reported loss.
Two Numbers From the Same Report
Intel closed its books on the second quarter of 2026 and released the results after markets shut on July 23[1][2]. Revenue was $16.1 billion, up 25% from a year earlier[1][2]. That's Intel's fastest quarterly growth in about 15 years[1]. Adjusted profit came in at 42 cents a share, nearly double the 21 cents Wall Street expected[2]. Shares jumped 11% to 13% in after-hours trading[6].
Further down the same report sat a different number: a net loss of about $11 billion, calculated under standard accounting rules known as GAAP[3]. Same company. Same three months. Two numbers that look like they can't both be true.
They are both true. They just measure different things, and the gap between them is where this quarter's real story lives.
Why a Rising Stock Price Shows Up as a Loss
In August 2025, the U.S. government converted about $8.9 billion in CHIPS Act grants into Intel stock, taking a stake of roughly 9.9% to 10% at about $20.47 a share[7][14]. Some of those shares sit in escrow rather than being handed over outright.
Accounting rules require Intel to re-value that escrow deal every quarter, based on Intel's current stock price. Picture it like a promise to hand over shares later: if the stock rises, that promise gets more expensive to keep, even though no cash actually moves. Intel's stock has climbed through 2026, so the value of that obligation climbed too.
Intel had to record that increase as a loss on paper. The bill came to about $12.5 billion, the main driver behind the $11 billion GAAP loss[3]. No money left the company. That's why Intel and most analysts treat the 42-cent adjusted profit as the truer read on how the business performed, while the GAAP loss is a real number that mostly reflects Intel's own stock doing well.
What Nobody Disputes
Set the accounting aside and the operating numbers aren't in question. Intel's Data Center and AI unit grew 59% year over year to about $6.3 billion, while Intel Foundry, the chip-manufacturing arm, brought in $5.8 billion, up 31%[5][8]. Adjusted gross margin was 41.8%, better than Intel's own guidance[2][8].
Intel also guided next quarter's revenue to $15.8 billion to $16.8 billion, above the roughly $15.1 billion analysts expected, and raised planned 2026 factory spending above $20 billion[2][8]. Much of that spending goes toward Intel's newest manufacturing process, called 18A, and its successor, 14A. Yields on 18A are reportedly running ahead of plan[8].
One gap stands out against all of that strength. Intel has not named a single large outside company as a customer for 18A, its most advanced process[3]. It did name Fortinet as a foundry customer, but for an older, less advanced process[3]. And before this report, Intel's stock had already fallen 28% in July, after climbing sharply earlier in the year[4].
The Bet That Isn't Settled Yet
That missing customer matters because Intel's real ambition goes beyond selling its own chips. It wants to become a contract manufacturer for other companies, a genuine rival to Taiwan's TSMC[3][8]. Strong data-center sales pay this quarter's bills, but Intel's stock price is riding on the foundry bet, and that bet still needs outside customers to prove itself[3][8].
That's also why some analysts call the stock priced for perfection. Intel trades at somewhere between 74 and 117 times its expected earnings for the next year, a measure called the forward price-to-earnings ratio[4]. That ratio is a rough gauge of how many years of today's profit it would take to earn back the stock's price, assuming profit stayed flat. Intel's 10-year average sits near 22; Nvidia currently trades under 20[4]. A ratio in the 70s or above means the market is pricing in years of fast, sustained growth, not just one strong quarter.
The government's stake raises the stakes again. By taking equity instead of just handing out grants, Washington made taxpayers co-owners of Intel's stock, for better or worse[7][10]. That turns Intel's share price into a political story: an asset for the administration if it keeps rising, a liability if it doesn't[10][9]. All of it also rests on continued heavy AI spending industry-wide. If that spending slows, the demand behind this quarter's growth would likely cool across chipmakers generally, not just at Intel[4].
Four Ways to Read the Same Quarter
Intel, led by CEO Lip-Bu Tan, frames this as proof its turnaround is showing up in results, not just promises. The company points to a seventh straight quarter of beating its own guidance and rising 18A output, arguing that AI demand and reshored U.S. manufacturing are paying off[8]. Intel has real reasons to push that framing: it needs new outside foundry customers to justify its enormous factory spending, and strong results help win them over[3][8].
The Trump administration treats Intel's rise as proof its industrial policy is working, pointing to a government stake bought at about $20.47 a share that is now worth far more on paper[8][10]. That case has support beyond the administration's own messaging. A Wall Street Journal investigation reported that the White House personally brokered commitments from Apple, Nvidia and SpaceX to work with Intel's foundry, and credited that effort with helping stabilize the company, while also noting that Intel Foundry still loses substantial money and that "political sponsorship creates serious risks if performance weakens"[15]. The administration has an obvious incentive to spotlight the win: a rising Intel stock is a visible, measurable result of its reindustrialization push[10][11].
Free-market critics counter that government ownership of a private company edges toward what they call "state capitalism." Their worry isn't that the deal was done improperly. It's that even a shareholder who says it will stay hands-off can still shape a company's hiring, pricing or factory decisions just by holding influence over it[9]. A CNBC poll found 49% of Americans said it's not appropriate for the government to own stakes in U.S. companies, against 19% who said it was appropriate[9]. These critics aren't challenging Intel's numbers; they're asking whether this ownership model should spread[9].
Wall Street skeptics don't challenge the quarter either. They challenge the price, pointing to the stretched forward P/E, the missing 18A customer, and an AI-spending cycle whose staying power is still unproven[3][4]. Seen from outside the U.S., particularly in Chinese-market coverage, Intel's revival looks less like a standalone win and more like one move in a broader U.S.-China contest over chip manufacturing. Outlets like the Hong Kong-based South China Morning Post keep the spotlight on Chinese chipmaker SMIC's own expansion instead[12].
The Coverage Split on Identical Facts
The range of headlines shows how much framing can shift a story built on the same set of numbers. Wire-style outlets, like Thailand's Kaohoon International, reported the quarter fairly straight, as a demand-driven beat with little editorial spin[2]. CNBC reported the beat alongside the loss and the missing 18A customer, giving weight to both sides of the ledger[2][3].
Bloomberg and Investing.com didn't dispute the numbers, but their headlines led with valuation risk, treating a record quarter as insufficient rather than as the story itself[4]. The White House's own release went furthest in one direction, crediting "America First" policy for the stock's rise while leaving out the $11 billion GAAP loss entirely[10]. The Wall Street Journal's reporting landed closer to the middle, independently crediting the administration's deal-brokering while still flagging that Intel Foundry remains unprofitable and that the strategy carries real political risk[15].
None of that coverage can yet answer the question hanging over the whole quarter: whether a major outside customer will actually sign on for Intel's most advanced chip process, and whether the AI spending fueling this rally still has room to run.
Summary
Intel reported second-quarter 2026 results after markets closed on July 23. Revenue was $16.1 billion, up 25% from a year earlier[1][2]. That was Intel's fastest quarterly growth in about 15 years[1]. Adjusted profit was 42 cents a share, roughly double the 21 cents analysts expected[2]. Sales and profit both beat forecasts, driven by strong data-center and AI chip demand[2][5]. Intel also guided higher for the next quarter[2]. Shares jumped about 11% to 13% in after-hours trading[6].
The same report showed a very different number lower down: a net loss of about $11 billion under standard accounting rules (GAAP)[3]. Intel said this came mostly from a roughly $12.5 billion non-cash charge tied to the U.S. government's stake in the company[3]. Here is the mechanism. In 2025 the government took an ownership stake in Intel as part of a CHIPS Act deal, with some shares held in escrow[7]. Accounting rules make Intel re-value that share arrangement every quarter. As Intel's stock soared this year, the value of that obligation rose, and Intel had to book the increase as a loss on paper[3]. No cash left the company. So bulls point to the adjusted profit as the 'real' operating result, while the headline GAAP loss reflects the stock's own rise. Both figures are accurate.
The genuine dispute is not over what Intel earned. It is over what the quarter proves. Supporters say Intel's turnaround is real: its new 18A manufacturing process is ramping, gross margins beat guidance, and data-center and AI revenue rose 59%[5][8]. Skeptics say the stock is priced for perfection, at about 74 times expected earnings versus a 10-year average near 22[4]. They note the stock fell about 28% in July before the report and that Intel still has not named a major outside customer for its most advanced process[3][4]. A second, political dispute runs alongside it: the Trump administration touts the government's Intel stake as vindication of its industrial policy, while free-market critics call government ownership of a private company a risky step toward 'state capitalism'[10][9].
The Event
On July 23, 2026, Intel released its financial results for the second quarter of 2026 (the three months ended in late June)[1][2]. The company reported revenue of $16.1 billion, up 25% year over year, and non-GAAP earnings of $0.42 per share, both above analyst consensus[2]. On a GAAP basis Intel reported a net loss of about $11 billion, which it attributed mainly to a roughly $12.5 billion non-cash mark-to-market charge related to shares held in escrow under its 2025 agreement with the U.S. government[3]. Intel shares rose about 11% to 13% in after-hours trading following the release[6].
Undisputed Facts
- Intel reported Q2 2026 revenue of $16.1 billion, up 25% year over year, its fastest quarterly growth in roughly 15 years[1][2].
- Non-GAAP (adjusted) earnings were $0.42 per share, versus consensus near $0.21; non-GAAP gross margin was 41.8%, above Intel's own guidance[2][8].
- On a GAAP basis Intel reported a net loss of about $11 billion, which it attributed mainly to a roughly $12.5 billion non-cash charge on escrowed shares tied to its CHIPS Act agreement with the U.S. government[3].
- Intel Foundry revenue was $5.8 billion, up 31%; Data Center and AI revenue was about $6.3 billion, up 59% year over year[5][8].
- Intel guided Q3 2026 revenue to $15.8 billion–$16.8 billion (16%–23% growth), above the roughly $15.1 billion consensus, and raised 2026 capital spending to more than $20 billion[2][8].
- Before the report, Intel shares had fallen about 28% in July, after rising sharply earlier in 2026; they jumped about 11%–13% after hours on the results[4][6].
- In August 2025 the U.S. government took an equity stake of about 9.9%–10% in Intel, converting roughly $8.9 billion of CHIPS Act grants into shares at about $20.47 each[7][14].
- As of the report, Intel had not named a major external customer for its leading-edge 18A process; Fortinet was named as a foundry customer using an older process[3].
The Pressure
Strip away the moralizing and blame. What structural realities persist regardless of which narrative wins?
- Prove the foundry, not just the quarter
- Intel's core bet is becoming a contract chipmaker that rivals TSMC. Data-center and CPU strength pays today's bills, but the stock's premium rests on landing large external 18A/14A customers. Until a marquee name signs, every beat is read as necessary but not sufficient[3][8].
- Industrial policy needs a poster child
- Washington converted CHIPS grants into equity to show taxpayers share the upside of reshoring. Intel's share price is now a political asset for the administration and a political liability if it falls[10][9].
- The AI capex cycle sets the ceiling
- Intel's data-center surge depends on continued heavy AI infrastructure spending. If that spending slows, the demand behind the beat and the guidance weakens across the sector, not just at Intel[4].
Material realityTwo accurate numbers describe the same company: an operating business that grew 25% with a 41.8% adjusted margin, and an ~$11 billion GAAP loss created almost entirely by a non-cash charge tied to the government stake as Intel's own stock rose[2][3][8]. Intel is spending over $20 billion this year on U.S. fabs and is ramping 18A, but its most advanced process still lacks a named large outside customer[3][8]. The stock is priced far above its historical average, so results that would be excellent for most firms can still disappoint a market expecting perfection[4].
Narrative as a weaponThree actors are shaping how you read this. Intel wants you to focus on the adjusted profit, the margin beat and 18A yields—the operating story. The White House wants you to see a soaring stock as proof its industrial policy works, and to look past the accounting loss it helped create. Wall Street bears want you to look past the record quarter and fixate on the valuation and the missing foundry customer. The dek of the original story ('shares extend 2026 rally on foundry strength') leans toward Intel's frame: the after-hours pop was real, but the stock had slumped 28% in July, and data-center/AI (+59%) was a bigger driver than foundry (+31%).
How Each Side Sees It
Each major actor’s view — how it frames things, its underlying incentive, and how it’s materially affected. Tap a side to read it.
Frames it asIntel argues the turnaround is now showing up in the numbers, not just in promises. It points to a seventh straight quarter of beating its own guidance, a non-GAAP gross margin of 41.8%, and 18A output rising with yields ahead of plan[8]. Its best case: leading-edge manufacturing is back on U.S. soil, AI is driving 'unprecedented demand for computing,' and raising capex above $20 billion signals real customer commitments for the 18A and next-generation 14A processes[8].
WhyIntel needs to win big outside 'foundry' customers to justify its huge factory spending and prove it can compete with TSMC. Strong results and confident guidance help it attract those customers, retain investors, and restore credibility after years of missteps[3][8].
Impact on themThe stock's direction affects Intel's cost of capital and its ability to fund U.S. fabs. A durable beat strengthens Tan's mandate; a stalled foundry ramp or a fading AI cycle would leave Intel with expensive, underused capacity[4][8].
Frames it asThe administration frames Intel's rise as proof that its 'America First' industrial policy works. By taking equity instead of just handing out grants, it argues, taxpayers now share in the upside of a reshored strategic industry. Officials highlight that the government's stake, bought near $20.47 a share, is now worth tens of billions on paper[10][8]. Independent reporting lends some support to this case beyond the administration's own messaging: a Wall Street Journal investigation found the White House personally brokered commitments from Apple, Nvidia and SpaceX to work with Intel's foundry, which it credits with helping stabilize the company[15].
WhyThe administration wants a visible win for its reindustrialization agenda and a template for taking stakes in other strategic companies. Intel's soaring share price gives it a political and fiscal talking point ahead of elections[10][11].
Impact on themThe government holds a large unrealized gain on its Intel shares[3]. But that same stake is what created Intel's $12.5 billion accounting charge, and it exposes the administration to blame if Intel later falters[3][9]. Even the Journal's more sympathetic account cautions that Intel Foundry still loses substantial money and that 'political sponsorship creates serious risks if performance weakens'[15].
Frames it asThis camp warns that a government ownership stake in a private company is a step toward 'state capitalism.' Their argument: even a 'passive' 10% federal shareholder distorts incentives, invites political pressure on hiring, pricing and factory locations, and blurs the line between the government as regulator and as owner[9]. They note that about half of Americans, in one poll, think it is not appropriate for the government to own U.S. companies[9].
WhyThese critics, including some economists and members of Congress, want to limit government intervention in markets and stop the Intel deal from becoming a repeatable model[9].
Impact on themIf the precedent spreads, it reshapes the relationship between Washington and major U.S. firms. If Intel's stock reverses, taxpayer money is directly exposed, strengthening their case[9].
Frames it asBears do not dispute the strong quarter; they dispute the price. Intel trades at roughly 74-117 times expected earnings depending on the estimate used, far above its 10-year average near 22, and by one measure over 100 times[4]. Nvidia, by contrast, trades under 20 times forward earnings[4]. Their crux: the AI spending boom may not last, chip stocks are already rolling over, and Intel still lacks a marquee external customer for 18A[3][4].
WhyInvestors and analysts are managing risk. After a huge run-up, they want to avoid overpaying for a company whose foundry payoff is still unproven[4].
Impact on themTheir caution helps explain the 28% July slump before earnings and keeps the stock volatile even after a beat[4][6].
Frames it asSeen from Beijing and much of Asia, Intel's revival is one move in a larger U.S.-China contest over advanced chips. Chinese-market coverage tends to fold the news into China's own drive for semiconductor self-sufficiency, highlighting SMIC's expansion and Beijing's push to use domestic equipment[12]. The implied argument: U.S. state backing of Intel mirrors, rather than contrasts with, China's own industrial policy[12].
WhyChinese firms and state media want to show that export controls have spurred, not stopped, domestic progress, and to downplay any single U.S. company's lead[12].
Impact on themA stronger Intel foundry tightens the U.S. lead in leading-edge nodes, raising the stakes for China's self-sufficiency targets and for firms like SMIC[12].
Like this article?
The Bias Ledger average rating 4
The same story, as framed by outlets across the spectrum, ordered least to most biased. The bias score (1 = straight, 10 = heavily spun) is an AI assessment of that framing — click an outlet to see its track record. The tell is the word choice or omission that reveals the angle.
| Outlet | Vantage | Bias | How they frame it | The tell |
|---|---|---|---|---|
| CNBC | U.S. center (business) | 2 | Intel reports fastest revenue growth since 2011, beats on earnings and guidance; notes $11B GAAP loss and missing foundry customer. | Reports the beat and the loss side by side, and flags the unnamed 18A customer. Straight but attentive to the caveats. |
| Kaohoon International | Thai (financial) | 2 | 'Intel Beats Earnings Expectations in 2Q26 as Data Center and AI Demand Continue Growing.' | Clean, demand-driven framing with little editorializing and little attention to the U.S. political angle. Reads as a wire-style market update. |
| The Wall Street Journal | U.S. center-right (business news) | 3 | 'The White House Made Fixing Intel Its Pet Project. It's Working.' | Independently credits the administration's brokered Apple/Nvidia/SpaceX commitments with stabilizing Intel, but balances this with the caveat that Foundry still loses substantial money and that 'political sponsorship creates serious risks if performance weakens' — steelmanning the industrial-policy case with reporting rather than promotion. |
| Bloomberg | U.S. center (financial) | 4 | 'Intel Needs More Than Blowout Earnings as Chips Rally Falters.' | The 'needs more than' framing subordinates a record quarter to a bearish thesis about valuation and the fading chip rally. Emphasis choice, not error. |
| Investing.com | U.S. center (markets analysis) | 4 | 'Intel Earnings Blowout Raises Questions Around a 117x Forward P/E.' | Concedes the 'blowout' but pivots immediately to valuation. The single most extreme P/E figure is chosen for the headline. |
| South China Morning Post | Hong Kong / Chinese-market (Alibaba-owned) | 4 | Frames the leading-edge chip race around China's SMIC and self-sufficiency drive rather than Intel's beat. | The omission is the tell: Intel's revival is contextualized as one front in a U.S.-China contest, keeping domestic Chinese progress in the foreground. |
| The White House | U.S. administration (Trump) | 9 | 'President Trump's Intel Deal and Soaring Markets Prove America First Works.' | A government release framing a company's stock gain as political vindication. Cites the taxpayer's paper profit; omits the GAAP loss and valuation risk entirely. |
References
- Intel Q2 Earnings: Revenue Up 25% to $16.1B — StockTitan · U.S. financial-news aggregator; market-data focused, low editorial slant
- Intel (INTC) earnings report Q2 2026 — CNBC · U.S. mainstream business news; center
- Intel Q2 2026 Earnings: Beat, $11B Loss, INTC Stock Reaction — INDmoney · Indian fintech/brokerage; retail-investor market coverage
- Intel Needs More Than Blowout Earnings as Chips Rally Falters — Bloomberg · U.S. financial news; center, markets-focused
- Intel Q2 Earnings Far Exceed Expectations: 59% Data Center Business Growth — TradingKey · Markets-analysis site; bullish/trading orientation
- Intel Shares Surge 13% in After-Hours Trading as Outlook Crushes Estimates — BigGo Finance · Taiwan-based financial aggregator; market-data focused
- Trump's 10% Intel Stake and the CHIPS Act equity conversion — Yahoo Finance · Aggregated U.S. financial news; center
- INTC Q2 2026 Earnings Call: CapEx Surge to Over $20B; margin 41.8% — BigGo Finance · Taiwan-based financial aggregator; earnings-call transcript summary
- Voters wary of government ownership in companies as Trump administration takes equity stakes — CNBC · U.S. mainstream business news; center
- President Trump's Intel Deal and Soaring Markets Prove America First Works — The White House · U.S. executive branch (Trump administration); official political messaging
- The US government's Intel stake is now worth $36 billion. Nobody in Washington planned it that way. — The Next Web · European tech media; center
- China's top chip foundry SMIC unveils action plan for seizing new growth opportunities — South China Morning Post · Hong Kong; Alibaba-owned; China-market perspective
- Intel Reports Second-Quarter 2026 Financial Results — Intel Corporation · Primary source; company investor relations
- Intel and Trump Administration Reach Historic Agreement to Accelerate American Technology and Manufacturing Leadership — Intel Corporation · Primary source; company/administration joint statement
- The White House Made Fixing Intel Its Pet Project. It's Working. — The Wall Street Journal · U.S. center-right; mainstream business/tech news