TD Cowen raises Applied Materials stock price target on WFE outlook

TD Cowen raised Applied Materials’ price target to $700 (from $525) and kept its rating despite the stock trading at $570.50 and appearing overvalued versus fair value. The company also launched six DRAM-oriented chipmaking systems plus an enhanced Centura Prime Epi (20% smaller footprint) to address the “memory wall” in AI; under a $250B WFE scenario, model earnings could approach ~$25 (vs. ~$13.35 FY26E). Offsetting sentiment, Michael Burry disclosed a short in Applied Materials and options activity was heavy (80,479 contracts; 49,633 calls vs. 30,846 puts), underscoring positioning-driven volatility.

Analysis

The market is rewarding the AI infrastructure trade, but the cleaner expression is in equipment names with direct exposure to process mix rather than the perceived AI kings. AMAT/KLAC/LRCX have more earnings leverage to a broadening capex cycle than NVDA, and the next leg likely comes from memory and advanced packaging, not just leading-edge logic. That makes AMAT the highest-beta beneficiary if customers keep pre-committing spend, but it also means the upside is front-loaded into expectations that can be repriced quickly.

The near-term risk is not a collapse in demand; it is a timing mismatch between share prices and actual order conversion. If the 2H26 WFE ramp slips even modestly, stocks that have already capitalized 2028 scenarios can de-rate 15-25% without any fundamental disaster. The take-or-pay angle reduces cancellation risk, but it also shifts the stress to customer balance sheets and makes the eventual downturn sharper if memory pricing weakens.

The broader contrarian read is that investors may be over-indexing on headline AI capex beneficiaries like CAT while underpricing the more direct semiconductor tool winners. Separately, the Starbucks-style build-vs-buy behavior is a warning for low-differentiation software vendors: enterprise customers are getting better at internalizing commodity workflows, which is a headwind for SaaS multiples but only a second-order issue for the platforms that provide the compute layer. In other words, the trade is not "AI everywhere"; it is "earnings visibility where the bottleneck sits."

Key Decisions for Investors

• Long AMAT / short NVDA, dollar-neutral, for 1-3 months: express the view that equipment names with direct WFE sensitivity can keep upgrading earnings while GPU leaders face digestion and multiple compression. Stop if NVDA resumes clear relative outperformance on hyperscaler spending commentary.
• Buy AMAT on a 3-5% pullback via a Jan-2027 650/800 call spread: defined-risk way to participate in the 2028 WFE upside without paying full equity multiple risk. Thesis breaks if AMAT loses its 50-day trend and order commentary fails to confirm the capex runway.
• Long AMAT / short LRCX for the next 1-2 quarters: prefer leading-edge/DRAM and packaging exposure over NAND-dependent convexity until NAND spending shows a real turn. Reverse the pair if NAND pricing and capex indicators inflect materially.
• Treat CAT as a crowded AI-infra beneficiary rather than a fresh long here; consider a tactical short only if data-center capex revisions slow, since valuation is already leaning on a continuation of the buildout narrative.