Why Intel's Foundry Revenue Isn't Keeping Up With Its Stock

Intel’s stock has jumped from about $19 to ~$120 over the past year, with market cap up nearly $500B, driven by revived CPU demand and renewed enthusiasm for Intel Foundry. However, foundry financials remain weak: Q1 2026 foundry revenue rose to $5.4B from $4.7B but external foundry revenue was only $174M, while the operating deficit was $2.4B (vs. $2.3B a year earlier), and 2025 external revenue totaled just $307M alongside a $10.3B operating loss. Intel is improving yields (estimated 50%-60% vs. TSMC’s 70%-80%), but the article argues evidence is still lacking—especially for large, sustained external production commitments at commercial scale.

Analysis

The market is still paying for an option on Intel’s foundry turnaround, but the cash economics do not yet support that premium. The key issue is not node capability in isolation; it is whether third-party customers will accept roadmap stability and process consistency from a supplier that has already changed course multiple times. Until external wafers become meaningful, the foundry remains a fixed-cost drag whose incremental improvements mostly reduce losses rather than create a high-margin growth engine.

Relative winners are the entrenched ecosystem leaders, especially TSM, which benefits from being the default benchmark for yield, scale, and execution. Intel’s progress can help hyperscalers like MSFT and AMZN negotiate better pricing and supply diversification, but that leverage is only valuable if Intel can convert evaluations into volume commitments; design kits and partnerships are not revenue. AAPL optionality is similarly overstated unless there is a disclosed manufacturing node and a production ramp, not just a feasibility check.

The contrarian risk is that investors are extrapolating “U.S. sovereign foundry” too far ahead of the operating data. Over the next 1-3 months, the main catalyst is not product launch hype but evidence of repeatable customer commitment and yield at commercial scale; absent that, the stock can re-rate down on disappointment even if the narrative stays positive. Over 6-18 months, the thesis is falsified if external revenue remains immaterial and foundry losses do not compress materially despite higher utilization.

Key Decisions for Investors

• Long TSM / short INTC as a 3-6 month relative-value pair; the trade works if execution credibility remains with TSM and Intel fails to show real external volume. Stop if Intel announces a multi-year foundry production commitment from a top-tier customer or external revenue inflects sharply.
• If forced to express downside in INTC, use a 3-6 month put spread rather than outright short stock; the stock can stay elevated on narrative alone, but the spread limits carry and benefits from any post-event fade. Falsifier: external foundry revenue above ~$500M/quarter on a sustained basis.
• Avoid chasing MSFT/AAPL on the headline; treat current disclosures as watch-items, not monetization. Reassess only if a named node and wafer-volume commitment are disclosed, which would convert strategic optionality into actual revenue.