The 6% drop after hours isn't about capex. It's about Susan Li telling the market 2027 is when this spend pays off, not 2026. $145B is loud. But Meta's ad revenue per DAU is up, and AI-driven targeting plus Reels recommendations are pulling more from the same user base. Capex/revenue near 70% looks insane only if you think Meta is GE in 2002. They aren't. They're funding the next ad-tech cycle while the rest of Mag 7 still pretends ROI is next quarter. The interesting tell is the related tickers on this topic: CRWV at $113.31 and AVGO at $406.94. CoreWeave didn't catch a bid AH despite being a direct beneficiary of hyperscaler GPU demand. That's the real signal. If CRWV and AVGO underperform into next week, the market is pricing AI infrastructure as commoditized, not strategic. That breaks
$Apple(AAPL)$ Ternus replacing Cook isn't the catalyst here. Apple's CEO transitions don't move fundamentals because the playbook is the company. Thursday's real read is whether Apple Intelligence has actual usage numbers behind it. Services is the floor again. 25% of revenue at fat margins, and the buyback prints earnings even when units wobble. iPhone units in China are the wildcard, but the comp is easy after the FY25 air pocket. What I'd watch on the call: any disclosed adoption metric for Apple Intelligence. Daily active users, queries per device, anything. 2 years in and the silence is louder than a miss would be. If Cook hands Ternus a clean Q2 with no AI numbers, the AI overhang just stays the AI overhang. $269.56 isn't a dip, it's f
$Intel(INTC)$ The bull case on Intel right now isn't the fabs. It's the chip in your old server rack. Xeon's quietly back in the mix because agentic AI still rides on CPUs, not just GPUs. Q1 made that obvious with data center up 22%. The headline keeps pointing at foundry, but that's not where the recovery's coming from. The fabs are still bleeding. External customers are basically nowhere. Yields are improving but that mostly helps Intel make Intel chips cheaper, which isn't a turnaround anyone should pay up for. Long INTC for the CPU side, not the foundry dream. You're eating fab losses either way, just be honest about which trade you're in.
$AST SpaceMobile, Inc.(ASTS)$ So Blue Origin's New Glenn dumped BlueBird 7 into the wrong orbit last Sunday. The upper-stage engine underperformed, the satellite couldn't get high enough to stay up, and now it's getting deorbited. One bird, gone. But the rest of the rollout hasn't even slowed. BlueBird 8, 9, and 10 are shipping in 30 days. AT&T's beta direct-to-cell service is still on for H1 2026. Verizon's already $100M deep into the partnership. The FY26 revenue guide of $150-200M (up from $70.9M last year) didn't move either. And Block 2 satellites carry 10x the capacity of Block 1, hitting 120 Mbps direct to phone. Constellation thesis is intact 🤷 The bigger story isn't ASTS though. It's launch risk. New Glenn just lost a cust
Look, the only reason I can hold RKLB through Neutron risk is the $1.85B backlog. Without that, $80 is too pricey for what's basically a launch schedule bet. Backlog grew 73% in a year. Most of it locks in revenue into 2028. So when RKLB drops 6% in a session like last Friday, I don't panic sell. The cashflow is coming whether or not Neutron flies on time. But Neutron is the actual upside. Stifel's $105 PT pencils in Q4 2026 first launch plus 10% medium-lift share by 2028 (against Falcon 9, sure). And it already got pushed back once from late 2025, dev costs creeping from $300M-ish to $360M. Confirm slip one more time lah, these things always do. The SpaceX IPO part is just sugar high. Filing buzz pops RKLB and DXYZ 3-5%, then fades. Nice tailwind, can't build a position on it. So $100 in