Stocks making the biggest moves after hours: Advanced Micro Devices, Super Micro Computer, Arista Networks & more

After-hours moves were driven by a mixed batch of earnings and guidance updates, led by AMD's 7% jump on second-quarter revenue guidance of $11.2B versus $10.52B expected and Super Micro's 19% surge on profit outlook above consensus. Negative reactions were also sharp, with Arista down nearly 14% on a slight gross margin miss and Lucid off 2% after a wider-than-expected loss and revenue miss. Several names added to the positive tone with beat-and-raise reports, including Jacobs Solutions and DaVita, while Klaviyo fell 18% on weaker-than-expected operating income guidance and a CFO departure.

Analysis

The market is rewarding execution where guidance implies demand is not just intact but accelerating, and it is punishing any sign that gross margin or revenue quality is slipping. The most important second-order signal is that AI infrastructure demand is still broadening, but the benefits are bifurcating: chip and server names with leverage to near-term capex are being re-rated higher, while networking and adjacent infrastructure names are being treated as margin-sensitive valves rather than pure growth assets. That sets up a rotation inside semis/infra rather than a clean sector-wide bid.

The AMD/SMCI reaction suggests investors are willing to pay for visibility into the next 1-2 quarters, but that also raises the bar for follow-through. If hyperscaler capex pauses even briefly, the high-beta beneficiaries can de-rate fast because the market has already pulled forward a lot of the good news; the vulnerable short leg is not necessarily the weakest balance sheet, but the name with the most expectation compression. Arista’s drop is the clearest example: the market is implying that even a modest margin miss matters more than line-item revenue, which can pressure peer valuation multiples if investors start assuming pricing or mix pressure is spreading.

Outside tech, the more durable positive surprise is in defensive/regulated cash generators that are quietly compounding through the cycle. That creates an interesting hedge against the AI/EV volatility trade: healthcare services and selective industrials can outperform if the market starts questioning whether high-growth hardware enthusiasm is sustainable. On the downside, EV and ad-tech exposed growth names are being marked as higher duration assets again, meaning they are likely to stay under pressure until management teams show either margin stabilization or a credible path to reacceleration over the next 1-2 quarters.

The contrarian takeaway is that the best risk/reward may be fading the strength in the most crowded beneficiaries and leaning into quality cash flow names where guidance revisions are less glamorous but more persistent. The consensus is likely underestimating how quickly sentiment can rotate from "AI spend is infinite" to "AI spend is concentrated and benchmark-sensitive," which would hit networking, secondary semiconductor suppliers, and software monetization stories first.

Key Decisions for Investors

• Reduce/avoid chasing AMD and SMCI strength after the gap-up; if long, use the next 1-3 sessions to trim into liquidity and tighten stops, since these names now trade on flawless follow-through and could give back 10-15% on any capex hesitation.
• Initiate a pair trade: long DVA / short KVYO over the next 2-6 weeks. The setup favors cash-flow visibility and defensive demand versus a software name facing margin pressure plus leadership transition; target a 8-12% relative spread with limited factor overlap.
• Buy ANET downside protection into the next 30-45 days rather than outright shorting immediately. The market is signaling that a modest gross margin miss can trigger multiple compression; a put spread captures downside if peer contagion spreads from networking margins to broader AI infrastructure suppliers.
• For opportunistic longs, buy J on dips as a lower-volatility beneficiary of capex/consulting resilience. Use a 3-6 month horizon and expect muted absolute upside, but better drawdown protection if the market rotates toward quality earnings and backlog visibility.
• Stay underweight LCID and SWKS until management proves stabilization over at least one more quarter. Both look like names where the market will require hard evidence before re-rating; upside is likely capped unless revenue inflections show up within the next 1-2 reporting cycles.