INTC · Intel Corporation · Stock research

Current Thesis

The foundry-revival re-rating that carried Intel from the $50s in April to roughly $110 in early July has been given back at speed. Intel is down about 31% month-to-date through 2026-07-16, versus TSMC's 15% decline over the same stretch, and the group backdrop is worse than the single name: SOXX has underperformed IGV by 19.95% in July, the widest monthly gap on record (2026-07-17), and the semiconductor ETF fell 13% in four weeks while erasing its entire premium to the Nasdaq 100 (2026-07-16). What broke was the multiple, not the estimates chip earnings forecasts are described as intact but a name trading on a 108x narrative multiple and +392% trailing-twelve-month gain is precisely the one that de-rates first when the group's premium compresses. The 2026-06-18 Truth Social domestic-manufacturing headline that produced the +8% session has still not been substantiated by any SEC-filed external foundry wafer-supply agreement. Q2 earnings land 2026-07-23, four trading days out, and that print now carries both the 18A yield question and the first real test of whether the July drawdown was multiple compression or the start of a fundamental reset. Entering ahead of it is a coin flip on a broken chart. The theme is MATURING at best and rolling toward SATURATED.

Bullish and bearish views on Intel Corporation

The model's bull view on Intel Corporation (INTC), in brief: 18A is shipping on High-NA, not just announced ASML confirmed Intel is already producing chips on its $400M EUV system (2026-07-15) and the two announced first high-volume logic product readiness on 18A (2026-07-15). The bear view: The drawdown is idiosyncratic, not just beta down ~31% in July against TSMC's ~15% (2026-07-16). Both cases follow in full.

Bull Case

• 18A is shipping on High-NA, not just announced ASML confirmed Intel is already producing chips on its $400M EUV system (2026-07-15) and the two announced first high-volume logic product readiness on 18A (2026-07-15). The process leg of the story has physical evidence behind it, which is more than most turnaround narratives carry.
• Sell-side targets still above spot Susquehanna held Neutral but raised its target to $115 (2026-07-16), stacked on Stifel $120 (2026-07-10) and Wells Fargo $110. Every one of those sits far above where the tape now trades, leaving room for a mean-reversion squeeze if the print clears.
• Enterprise-AI partnership widened Intel expanded its Google Cloud relationship and rolled out Gemini Enterprise for AI-assisted chip design, publicly endorsed by Sundar Pichai (2026-07-17). Small revenue impact, real credibility signal on design-cycle speed.
• Group valuation reset creates a floor argument with Micron at 6.8x forward earnings and SOXX's Nasdaq-100 premium gone (2026-07-16), the semis complex is no longer priced for perfection; a rebound in the basket lifts Intel mechanically through SMH/SOXX flow.
• Loud retail sponsorship persists Cramer named Intel his No. 1 stock on 2026-07-15 and tied it to the ASML capex cycle. Sponsorship of that volume is a two-way signal but it does keep bid depth on dips.

Bear Case

• The drawdown is idiosyncratic, not just beta down ~31% in July against TSMC's ~15% (2026-07-16). Intel is losing twice the group. The most-extended winner unwinding hardest is exactly how a crowded re-rating ends.
• Foundry economics remain the unanswered question TSMC printed a 67.7% gross margin with an accelerating 2nm ramp and a $64B capex outlook (2026-07-16/17). Intel Foundry is nowhere near those economics, and TSMC's spending guide raises the bar Intel must clear rather than lowering it.
• JPMorgan's Q3 top-short call (2026-07-10) has already worked. That thesis was anchored on extension; with the extension partly worked off, the risk shifts to whether the desk's fundamental leg foundry cash burn gets confirmed on 2026-07-23.
• The political catalyst never converted no filed wafer agreement, no named external customer, no node or volume commitment behind the 2026-06-18 headline. Headline-driven re-ratings that go unconfirmed for five weeks tend to fully retrace.
• x86 attacked from the AI-PC side NVIDIA's ARM-based AI-PC silicon (2026-06-01) targets the client CPU franchise that funds the foundry build. Q2 client CPU share is the number that matters on the print.
• Earnings four days out a binary on a chart that has already lost its structure. Position risk here is gap risk, not thesis risk.

Setup & Price Structure

The post-2026-06-18 announcement shelf near $92 has been lost, and with it the entire structural argument for the June leg. A roughly 31% July decline from the ~$110 area puts the tape in the mid-$70s, well below the analyst target band of $110–$120 and materially below the golden-cross structure that defined the spring advance. The stock fell 3% premarket on 2026-07-16 on TSMC capex worries alone, which shows how little independent bid remains. What was a stretched-above-the-200DMA problem in early July is now a broken-uptrend problem: the name needs to build a base and reclaim prior support before any technical entry exists. Buying this level ahead of a print because it is 30% cheaper than three weeks ago is averaging into a broken structure with a calendar bomb attached. Wait for the print, then judge the reaction candle.

Catalyst Calendar (next 30 days)

• 2026-07-23 Q2 2026 earnings. The binary: 18A yield commentary, Intel Foundry operating loss trajectory, client CPU share versus AMD, and any named external foundry customer. Flagged as a key S&P 500 name to watch next week (2026-07-18).
• ~2026-07-24 to 2026-07-31 (est.) post-print analyst target revisions. Susquehanna at $115 and Stifel at $120 are stale against spot; a cluster of cuts confirms the de-rate, a cluster of reiterations marks the multiple reset as complete.

Elapsed catalysts

• ~2026-08 (est.) follow-through 18A / High-NA milestone disclosures via ASML's supply-chain commentary, after the 2026-07-15 first-high-volume-logic readiness announcement. _(passed 4d ago)_
• Ongoing Taiwan-risk headlines. Pat Gelsinger's 2026-07-17 warning on a China-induced Taiwan power interruption is the tail scenario that would re-rate Western fabs violently upward; unpredictable, not investable as a base case. _(passed 2d ago)_

What Would Change Our Mind

An 18A external customer disclosed by name in an SEC filing with node, volume and timeline would convert this from a headline story into a contracted one, and would justify paying a premium multiple again. Second, a Q2 print on 2026-07-23 showing Intel Foundry's operating loss narrowing faster than guided alongside flat-to-up client CPU share, followed by a weekly close reclaiming the $92 shelf, would mark the July decline as multiple compression that has run its course. Absent both, the read stays negative: a weekly close below $70 completes the full round-trip of the June announcement move and confirms the re-rating was a political headline priced as a contract.

Correlation Notes

• Do not stack with DELL or HPQ shared AI-PC refresh exposure plus the same NVIDIA-ARM x86 threat. One position maximum across that cluster.
• SOXX/SMH beta dominates the single name right now. With the semiconductor complex posting its worst month versus software on record (2026-07-17), Intel will not decouple upward while the basket is being sold. Group direction is the first-order variable.
• TSMC is the read-through, and it is a negative one. TSMC's 67.7% margins and $64B capex frame the competitive gap; TSM weakness transmits to INTC with amplification, as 2026-07-16 showed.
• ASML is the cleaner expression of the same equipment cycle it rallied on earnings and a capacity expansion (2026-07-15) while Intel fell, meaning the tool vendor captures the capex narrative without carrying the execution risk.