Intel Earnings Beat Fizzles as Stock Slides Back to $20 Support

Intel’s Q2 2025 report was a tale of two narratives: optimism around CEO Lip-Bu Tan’s decisive cost-cutting and restructuring strategy, and ongoing market skepticism about Intel’s long-term turnaround. Shares initially jumped on better-than-expected revenue and commentary around lower CapEx, but reversed course during the call, ultimately retesting the $20 support level—an area that’s held since May. The initial pop was short-lived as investors weighed weak gross margins, a mixed Q3 outlook, and persistent structural challenges.

WATCH: Bob Elliott: Markets Are Delusional — And Credit Knows It

Intel reported Q2 revenue of $12.9 billion, beating estimates of $11.92 billion and exceeding its own guidance. However, adjusted EPS came in at a loss of $0.10, below the $0.01 expected. Gross margin was a glaring weak point, printing at 29.7% on a non-GAAP basis, versus the 36.5% consensus. Intel attributed much of the margin compression to one-time restructuring and impairment charges, including $1.9 billion related to layoffs and factory write-downs. Stripping out these effects, management claimed margins would have been 37.5% and EPS positive—though the market didn’t bite on that adjustment.

Tan’s restructuring efforts are aggressive and immediate. Intel will end the year with a core workforce of roughly 75,000 employees, following a 15% headcount reduction. Management also scrapped major fab projects in Germany and Poland, consolidated assembly operations from Costa Rica into sites in Vietnam and Malaysia, and slowed the Ohio fab construction. The goal: realign spending with actual demand rather than speculative future growth. Tan described the prior footprint as “needlessly fragmented and underutilized” and emphasized a more disciplined CapEx strategy going forward.

For 2025, Intel reaffirmed CapEx at $18 billion gross, with $8–11 billion net after CHIPS Act offsets, and suggested 2026 would be lower. Operating expenses are targeted at $17 billion in 2025, declining to $16 billion in 2026. These are big steps for a company long criticized for throwing capital at problems without securing demand. In that context, the market initially rewarded the shift—but as the earnings call wore on, questions resurfaced about whether leaner spending alone can offset margin pressure and competitive losses.

Q3 guidance added more uncertainty. Intel expects revenue between $12.6 and $13.6 billion, roughly in line with consensus, but sees breakeven EPS versus expectations for $0.04. Non-GAAP gross margin is projected at 36%, slightly below Street forecasts. Lunar Lake ramp costs and early Panther Lake development are expected to weigh on margins near-term. CFO David Zinsner offered reassurance that Panther Lake, launching later this year, could help margins rebound in 2026 as volumes increase.

Tan didn’t shy away from hard truths. “We approached AI with a traditional silicon and training-centric mindset,” he said, outlining a pivot toward a full-stack AI strategy focused on inference and agentic AI. Intel also admitted it had overbuilt capacity in the past, and future investments—particularly around the 14A process—would be tied to confirmed customer commitments. That’s a change in tone from the aspirational language of prior years, replaced with a more grounded, trust-building approach.

The foundry business remains a question mark. Intel Foundry posted Q2 revenue of $4.4 billion but an operating loss of $3.2 billion, weighed down by $800 million in impairments. Tan reaffirmed that 18A remains on track, with Panther Lake ramping by year-end, and framed the process node as the foundation for three product generations. But he was clear: 14A won’t move forward without outside interest. Analysts pressed for specifics, and Tan hedged—saying customer engagement was improving but volumes had yet to materialize.

Intel’s Client Computing Group posted $7.9 billion in revenue, up sequentially from Q1 but down from $8.1 billion a year ago. That was supported by PC refresh tailwinds, including Windows 10 end-of-life upgrades and some customers front-loading purchases ahead of tariff risks. Tan acknowledged both dynamics but warned that without 18A success, that strength may not sustain into Q4.

Market sentiment remains mixed. Bulls see Tan’s bold approach as a necessary correction after years of bloat and strategic drift. Bears point to deeper competitive issues—especially in server CPUs, where Intel continues to lose share to AMD and ARM. And while 18A could become a catalyst, it hasn’t launched yet. For now, the stock remains range-bound, with a “prove it” valuation under $100 billion and few near-term catalysts beyond Tan’s playbook.

In the end, Intel’s Q2 results show a company in transition. Revenue surprised to the upside, but earnings and margins remain challenged. Cost-cutting is well underway, and Tan’s actions suggest a willingness to break with Intel’s old ways. But Wall Street isn’t awarding style points just yet. Investors want execution—on foundry, AI, and 18A—and until then, rallies may continue to fade as quickly as they appear.