Shares of Intel and AMD Have Plummeted in the Last Few Days. Is It Time to Panic?

The article highlights a sharp July sell-off: AMD is down more than 10% and Intel down about 20% after strong 2026 runs (AMD ~+140%, Intel ~tripled). It argues AMD’s data center growth of +57% YoY to $5.8B remains far behind Nvidia’s +92% YoY to $75.2B, and Intel’s foundry revenue rose 16% YoY to $5.4B but still lagged TSMC’s $35.9B (+41% YoY). The key bearish point is valuation—AMD trades ~70x forward earnings vs Nvidia ~22.8x, and Intel ~100x vs TSMC ~27.5x—implying AMD/Intel must triple/fourfold earnings beyond 2026 projections to justify prices.

Analysis

The market is still paying for optionality rather than execution, and this note reinforces a likely near-term unwind in the two highest-beta turnaround names. The cleanest expression is not outright bearishness on semis, but a relative reallocation from monetization laggards to proven cash compounding: NVDA and TSM should continue to absorb AI capex even if the broader semiconductor tape stays firm. That implies AMD/INTC can underperform on multiple compression even if their businesses keep growing, because the market is increasingly unwilling to pay “future winner” premiums without visible share gains.

The second-order effect is positioning. After a strong first-half move, both names are vulnerable to factor rotation out of momentum/turnaround baskets and into quality-growth. That can spill into passive and ETF flows: SOXX/SMH may hold up, but their internal leadership should skew toward NVDA/TSM, with AMD/INTC acting as source of funds on any sector bounce. For Intel specifically, the foundry story only matters if it starts to alter external customer behavior; otherwise, it remains a long-dated promise that competes for capital against a far better-financed incumbent.

Contrarian risk: the consensus is treating the gap between these companies and their leaders as static. If AMD’s data-center mix improves on inference workloads or Intel lands a credible foundry anchor customer, the market could rapidly re-rate the path-to-profitability and squeeze shorts. The tell will be forward EPS revisions and gross margin trajectory over the next 1-2 quarters; absent that, the current multiples look like they need another reset rather than a fresh leg higher.

Key Decisions for Investors

• Over the next 1-3 months, favor a relative-value pair: long NVDA / TSM, short AMD / INTC in equal dollar weights. Thesis: AI capex and foundry share continue to concentrate in the winners while turnaround multiples compress; risk/reward is better than chasing outright index exposure.
• Use any 2-3 day rebound in AMD to trim or hedge exposure rather than averaging down. If AMD fails to show a meaningful step-up in data-center gross profit or forward EPS revisions into the next earnings cycle, the stock remains a candidate for another leg lower.
• For investors needing semis beta, rotate out of AMD/INTC into SMH or NVDA call spreads instead of naked longs in the turnarounds. This keeps participation in the AI theme while reducing exposure to valuation de-rating.
• Set a watch item on Intel foundry: if external customer wins do not materialize within 1-2 quarters, treat the foundry premium as narrative-only. A negative reset there would justify staying short INTC versus TSM.
• Falsifier for the bearish relative view: AMD data-center growth reaccelerates materially versus Nvidia on a revenue base that narrows the gap, or Intel posts clear external foundry bookings that change 2026-2027 earnings power.