AMD — long thesis

Confidence: medium · Status: active

Thesis

AMD delivered a blowout Q1 2026: revenue $10.25B (vs $9.89B expected), EPS $1.37 (vs $1.29), and Q2 guidance of $11.2B implies 46% YoY growth — well above Street expectations. Data center revenue surged 57% YoY to $5.8B. Both OpenAI and Meta have signed up for AMD's Helios rack-scale AI systems (shipping H2 2026). Meta committed to a multiyear deal deploying up to 6 GW of AMD GPUs. The AMD + Intel x86 AI Compute Extensions collaboration adds a CPU moat. However, the stock has already rallied ~66% YTD and ~253% over the past year, trading at ~40x forward earnings. HSBC downgraded to Hold citing stretched valuation, and consensus PT of $307.50 sits well below the current price. This is a high-quality company but valuation risk is elevated — entry should be on a meaningful pullback from post-earnings euphoria levels, not at the gap-up.

Triggers

Entry: Enter on a pullback of 10-15% from post-earnings high. If stock pulls back to the $330-340 range, consider a small position (~1-2% of portfolio). Avoid chasing the 16% post-earnings gap-up. The HSBC fair value of $340 and consensus PT of $307 suggest limited upside at current levels. Wait for consolidation.

Exit: If entered on pullback, take profits at 15-20% gain from entry. Exit if Q2 2026 earnings show data center growth decelerating below 40% YoY or if Q2 guide disappoints. Trim if position exceeds 3% of portfolio.

Invalidation: MI350/Helios shipments delayed or major customer (Meta, OpenAI) reduces orders; data center revenue growth drops below 40% YoY for two consecutive quarters; stock breaks below $280 (pre-earnings level) on fundamental deterioration; TSMC capacity constraints materially limit AMD's ability to ship.

Cited evidence

News