NVDA's CPO milestone could be the optics tailwind: Nvidia says co-packaged optics just entered mass production. There's been a ton of FUD on how fast CPO actually ships, so this is at least a real data point instead of a roadmap slide. If it holds, that's a tailwind for the photonics names already on the radar, watching for follow-through, not chasing the headline.
Third firm now, $200B+ WFE forecasts for 2027 keep stacking up: Another shop's out with a $200B+ wafer fab equipment forecast for 2027, the third one now, with $LRCX, $AMAT and $KLAC the main beneficiaries. TSMC's aggressive N2/N3 capacity buildout is the driver behind the raised capex numbers. Adds more weight to the WFE thesis already playing out in $LRCX's own print.
UBS says memory could eat 73% of hyperscaler capex in 2027: UBS is calling for hyperscalers to spend $761B, or 73% of total capex, on memory alone in 2027. For context, total hyperscaler capex this year is around $760B, so that's the entire current spend level going just to memory next year. Big number, sustainability's an open question, but it's another data point for how far the memory supercycle thesis could run.
PENG's memory-peer comp doesn't really hold up: $PENG gets lumped in and traded with the memory names, but the read here is that's not the right comparable even as memory ASPs expand. Worth a second look at what's actually driving the multiple before assuming it moves 1:1 with the memory complex.
Borrowing O'Neil's follow-through day playbook to read this bottom: Leaning on William O'Neil's old framework for spotting a durable bottom, a rally attempt has to hold a low then close up from it, then you wait for a real follow-through day to confirm before trusting it. Useful extra lens right now given how many false bounces we've already sat through. If the market can't produce a clean follow-through day, treat any bounce as noise, not the bottom.
Sitting on buying power until the bottom actually confirms: No new stock recs for now, the call is preserve buying power while the market's still carving out a bottom rather than force entries. A watchlist of names on the radar is coming tomorrow for anyone sitting on cash they want to deploy, but for now it's mostly about tracking existing portfolio names, not adding new ones.
Earnings calendar thins out, ORCL and NVDA the ones left to watch: Most of the mega caps are done reporting, this week's lighter, with $ORCL and $NVDA the next real catalysts further out. Not much scheduled to move the tape near-term, so no reason to force anything into an empty calendar.
Aggregated and distilled by Two Sentence Traders. Not financial advice.
