Big Tech and its silicon suppliers just turned the after-hours tape into a stress test. Meta’s earnings and a $115 billion to $135 billion 2026 capex plan lit up anything tied to AI infrastructure, while ad-tech comps and mega-cap defensives chopped on the headline crossfire. The sector with the steering wheel tonight was technology, specifically the AI buildout complex: hyperscale platforms, ad engines, and the chipmakers who sell them shovels.
Meta delivered the quarter and the manifesto. Revenue around $59.9 billion and EPS of $8.88 gave the Street its beat, but the real show was 2026: capex of $115 billion to $135 billion, expenses tracking toward $169 billion, and a pledge that operating income in 2026 will top 2025 even after the spend ramp. Shares spiked roughly 10% on the print, then bled to about a 2% after-hours decline as capex sticker shock replaced euphoria. What drove attention: a deliberate pivot from “social app” to industrial buildout, with Meta Compute, third-party cloud bridging, and data center expansion underscoring that AI’s bottleneck is now measured in gigawatts and depreciation. Trading profile: extension-session volatility, heavy liquidity, and a price action round-trip that tells you the market wants the AI capacity story but will tax it on margins. Key takeaway: AI is shorthand for capex, and Meta has the ad cash flows to fund it. Ad impressions rose 18% year over year with average price per ad up 6%, and 3.58 billion daily active family users is a moat you can lever. But operating margin just slipped to 41% from 48% a year earlier and R&D hit $17.1 billion in Q4; debt also climbed to about $58.7 billion after a big Q4 issuance. Translation: the company is asking to be graded like an infrastructure play with an ad-tech engine. If the engine hums and guidance stays buoyant, the market will tolerate the build.
When the hyperscalers announce spend, the tape hunts for suppliers. Intel vaulted into the activity leaderboard with a volume surge and a price pop north of 6%, closing around $25, according to tape checks. What drove attention: investors leaning into AI infrastructure exposure across CPUs, networking, and packaging as hyperscale capex resets higher. The story for Intel remains simple and hard at the same time: win sockets, ship silicon that feeds training and inference clusters, and reboot foundry credibility while the industry needs power-efficient compute and interconnect at scale. Trading profile: outsized volume, high beta to the AI cycle, and a stock that’s been a laggard enough to attract rotation when the theme catches heat. Key takeaway: the order book for AI buildouts is stretching, and even a partial share of that wallet can be a multi-year lift. But Intel still has to execute on product roadmaps and foundry timelines in a market that won’t wait. The opportunity is real; the tolerance for missteps is not.
AI turns memory from commodity to choke point, and the tape keeps rewarding that reality. Micron climbed about 2% after hours to roughly $446, riding the same hyperscale capex wave that Meta just supercharged. What drove attention: investors pricing tighter supply and structurally higher DRAM and high-bandwidth memory content per server as training and inference footprints expand. The industry backdrop favors disciplined supply and richer mix, which magnifies operating leverage when pricing tails the cycle higher. Trading profile: steady bid on AI headlines, with moves that tend to extend when hyperscalers specify bigger budgets and more racks. Key takeaway: as long as AI servers are multiplying, memory bits and bandwidth are revenue per rack. Micron is levered to that mix shift; execution in HBM and next-gen DRAM nodes is the swing factor. In an AI capex upcycle, memory vendors look less like commodity suppliers and more like toll collectors.
Apple drifted about 0.3% lower after hours to roughly $258. What drove attention: less about a fresh headline, more about positioning as the market digested Meta’s spend-now narrative and rotated through tech megacaps. Apple remains the cash-flow machine that sells premium hardware and services to a captive base, with a developing on-device AI strategy that is more consumer-facing and less data-center-levered. That makes it less immediately sensitive to a hyperscale capex surge, and that’s exactly how the tape treated it tonight. Trading profile: low-drama, narrow move, liquidity always there, implied volatility tame relative to AI beta names. Key takeaway: Apple won’t live or die on Meta’s capex number. The stock’s next real catalysts are product cycle durability, services growth, and whether on-device intelligence can reignite upgrade behavior without crushing gross margins. It’s a steady compounder caught in a night owned by infrastructure spend.
Alphabet eased about 0.9% in the after-hours session, a sympathy sag as investors triangulated ad-tech comps and the cost of chasing AI at scale. What drove attention: tonight was Meta’s party, but Alphabet is in the same club, threading the needle between AI investment and ad monetization resilience. Context matters: in 2025, Alphabet posted 60% plus gains as the company proved AI could strengthen, not cannibalize, core search and advertising, with AI-powered results reportedly boosting engagement, ad pricing, and margins. The Gemini 3 release flipped the narrative last year; now the Street wants to see that momentum without a runaway cost line. Trading profile: modest slip on a busy tape, underperformance to semi beneficiaries, and a valuation that already discounts a lot of AI competence. Key takeaway: Alphabet’s advantage is distribution scale and an ad stack that handles intent like no one else. The risk is letting infrastructure spend outrun visible returns. Expect tighter scrutiny on unit economics for AI features inside Search, YouTube, and Cloud.
Technology was the most active sector because hyperscalers just told you where the money is going: concrete, chips, and power for AI. That spend cascades through semis first, then filters into platforms that can monetize the experience without detonating margins. The trade is simple, not easy: own the capacity builders with cash fountains or the suppliers with pricing power, and fade the stories that need perfect execution to justify perfect multiples.
