European stocks subdued amid Nvidia results, Iran war developments

Nvidia posted record quarterly sales of $81.6 billion, up 85% year over year, and net income of $58.3 billion, more than triple the prior year, both ahead of Wall Street estimates. CEO Jensen Huang said demand for the company's AI systems is turning 'parabolic' as the market enters the 'era of agentic AI,' though the stock edged lower after hours because the outlook excluded China sales and was only modestly above expectations. The strong results are supportive for the AI theme and broader chip sentiment, even if elevated expectations temper the immediate stock reaction.

Analysis

The market is rewarding the broader AI complex, but the more interesting signal is that Nvidia continues to print upside without clearing the bar for an outright re-rating. That usually means the “good news” is being arbitraged into positioning rather than fundamentals, so the first-order move can still be positive while forward returns compress as expectations reset higher again. The bigger takeaway is that capital expenditure momentum around AI infrastructure remains intact, which supports the entire equipment-to-power chain even if NVDA itself trades more like a crowded momentum asset than a clean earnings compounder.

Second-order winners are likely to be the less-obvious beneficiaries of another year of buildout: foundry equipment, advanced packaging, networking, memory, and power/density bottlenecks. If agentic AI is the next narrative leg, the bottleneck shifts from model training to inference at scale, which favors systems that optimize throughput, latency, and energy efficiency rather than just GPU count. That argues for relative performance in names levered to rack-level integration and data-center power delivery, while pure-play semiconductor beta may remain hostage to sentiment swings after every earnings print.

The main risk is not a demand rollover over the next few weeks; it is a classic “expectations inflation” trap over the next 1-3 months, where strong results fail to produce follow-through because investors were already positioned for perfection. The China exclusion in guidance also matters more for the ecosystem than for NVDA near term: it leaves a latent upside option for the supply chain if export constraints ease, but it also keeps headline estimates artificially clean and raises the probability of disappointment when the next revision cycle arrives. If the stock can’t hold post-earnings gains despite record fundamentals, that is usually a warning that the trade is becoming crowded.

The contrarian setup is to fade outright NVDA upside while staying constructive on the enablers that benefit from capex durability. In other words, the market may be underappreciating how much of the AI spend wave has already been socialized into mega-cap earnings expectations, but still underpricing the slower, steadier winners one layer down the stack. That creates a better risk/reward in relative-value expressions than in chasing the headline leader after a blowout quarter.

Key Decisions for Investors

• Short-term: sell upside calls / initiate a call spread overwrite in NVDA into strength over the next 1-2 weeks; the setup favors limited further multiple expansion with elevated post-earnings positioning risk.
• Relative value: long SMH basket vs short NVDA for 1-3 months; express the view that the broader AI supply chain can keep compounding even if the market stops rewarding the single-name bellwether.
• Overweight AI infrastructure enablers with better expectation gaps than NVDA over the next quarter, especially networking and power/density beneficiaries; target names that can rerate on capex durability rather than headline earnings beats.
• If NVDA retraces to pre-earnings breakout levels within 2-4 weeks, buy the dip only via call spreads, not outright stock, to keep theta and multiple-compression risk contained.
• Watch for a 4-6 week drift lower in post-earnings momentum; if the stock underperforms while the semis complex stays firm, rotate capital from NVDA into second-order winners and treat that divergence as confirmation the consensus has become too one-sided.