Merck Q1 earnings fall but beat expectations; 2026 guidance raised

Merck reported Q1 net profit of 669 million euros, down 9.4% year on year but ahead of expectations for 2.11 euros per share versus 1.99 euros. Net sales fell 2.8% to 5.13 billion euros, also topping estimates, and the company cited strong demand in Electronics tied to AI and high-performance computing materials. Merck slightly raised 2026 EBITDA guidance to 5.7 billion-6.1 billion euros from 5.5 billion-6.0 billion euros and reaffirmed 2026 sales of 20.4 billion-21.4 billion euros.

Analysis

The key signal here is not the headline earnings beat; it is that the semis-related portion of Merck’s portfolio is still growing into a backdrop of cyclical softness. That matters for NVDA because materials demand is a cleaner upstream read-through than finished-chip demand: when wafer starts, advanced packaging, and AI/HPC node intensity stay firm, suppliers with content in lithography, etch, and process materials tend to see order durability even if headline capex slows. In other words, the market may be underestimating how sticky AI infrastructure spending can be once platform buildouts move from pilot to scaling.

The second-order effect is competitive pressure on the broader semiconductor supply chain. If Merck is seeing strength from advanced applications, adjacent names that depend on continued node migration and packaging complexity should benefit, while lower-end mature-node exposed suppliers remain at risk of mix degradation. This also argues that AI trade leadership may broaden from pure compute to picks-and-shovels enablers over the next 1-3 quarters, especially if hyperscalers keep capex elevated but become more selective on hardware procurement.

The main risk is that this is a lagging indicator: materials demand can stay resilient for several quarters after customers have already started trimming forward orders. The reversal catalyst would be any capex deferral from the top-tier foundries or a sharp inventory correction in the semiconductor channel, which would show up first in bookings rather than revenue. For NVDA specifically, the setup is supportive but not explosive; the more asymmetric expression may be in suppliers with operating leverage to AI/HPC mix, where consensus still prices a normal cyclical slowdown.

Contrarian view: the market may be over-focusing on whether AI demand is "real" and underappreciating the time lag between infrastructure digestion and supply-chain normalization. If that lag persists, the winners are not just the obvious GPU names but the enabling ecosystem, while any disappointment in near-term chip revenue could create a temporary mispricing opportunity in the broader semiconductor complex.

Key Decisions for Investors

• Add to NVDA on pullbacks over the next 2-6 weeks, but size it as a continuation trade rather than a breakout trade; the read-through is supportive for AI demand durability, not a catalyst for immediate multiple expansion.
• Initiate a basket long in semiconductor equipment/materials beneficiaries for 1-3 months; pair against a short in mature-node/commodity semiconductor exposure to express the mix shift implied by advanced-node demand.
• Use call spreads in NVDA into the next 1-2 earnings windows instead of outright calls; upside is supported by supply-chain confirmation, but implied volatility likely overstates near-term upside.
• Watch for any foundry or hyperscaler capex commentary over the next 4-8 weeks; if bookings weaken before revenue does, reduce exposure quickly because materials strength will be a lagging indicator.