everyone calls the MU drop profit taking but ~18% in a month is a real move. it crossed 1T market cap in may and now people act like nothing happened. holding my core but not adding another share until i see the next HBM print
everyone calls the MU drop profit taking but ~18% in a month is a real move. it crossed 1T market cap in may and now people act like nothing happened. holding my core but not adding another share until i see the next HBM print
$Taiwan Semiconductor(TSM.US) posts a blowout, raises guidance, and the sector sells off because it also raised capex to 60-64B. let me get this straight, the market's mad the foundry is investing MORE to meet demand? that's backwards. TSMC doesn't overbuild, ever. a capex raise this big means the order book is screaming. i hold TSM and this reinforces the whole AI thesis for me, not weakens it.
lost in the memory dump: $ASML(ASML.US) crushed Q2, sales up 21% to 9.3 billion euros, and raised its 2026 guidance, stock up 7%. ASML is the monopoly on the machines that make every advanced chip, so when it raises guidance, it's saying AI capex is still accelerating, not slowing. that's the bullish counterpoint to the memory selloff. the buildout is funded. i'm long the equipment side as the cleanest capex proxy.
the $NVIDIA(NVDA.US) China headline cuts the other way from how people read it. the "white list" halving Asian buyers sounds bad, but Nvidia already wrote China DC revenue to zero last quarter, so any approved shipments now are upside, not downside. at a new high with the bad news already in the numbers, the setup is asymmetric. holding, adding only on a market flush.
everyone's staring at tonight's CPI but the print i actually care about is $Taiwan Semiconductor(TSM.US) on Thursday. it's the AI chip bellwether, it makes the silicon for Nvidia, Apple and AMD, so its guidance is the least-biased demand read in the whole trade. after this week's memory scare, a strong TSMC guide would basically end the "AI is topping" argument. holding into it, not adding before earnings.
the thing keeping me long $NVIDIA(NVDA.US) is that every catalyst keeps confirming the demand: SK Hynix debut, Meta doubling compute, neoclouds expanding. Jensen says demand isn't slowing and the memory makers agree it's undersupplied till 2030. at a market-average multiple that's a rare combo of leader plus reasonable price. holding, adding only on a real flush.
the thing about $NVIDIA(NVDA.US) at 203 is it's below its ATH and trades around 21x forward earnings, market-average multiple for the most important AI company on earth. Meta wants to double compute to 14GW next year, every capex headline is Nvidia demand. i added a small piece around 190 last week and i'm holding the lot. not chasing, just letting it compound.
the wild part about $NVIDIA(NVDA.US) at 204 is the valuation. forward P/E around 21.7, basically the market average, for the most important company in AI. meanwhile the names getting dumped trade at triple-digit multiples. cheap and winning at the same time. i added a small piece last week around 190 and i'm holding the lot, letting the demand signals keep stacking.
$NVIDIA(NVDA.US) "roadmap intact" saved my position Monday morning 😮💨 i panic-trimmed a piece last week on the Kyber delay rumor and bought it back when the denial came. net result: paid spreads for nothing. lesson: do not trade off semianalysis leaks on Nvidia. holding the full position now, not touching it.
UOB sitting at an all-time high, up about 17% year to date and it beat estimates last quarter. the whole STI bank trade has been the quiet winner while everyone stares at US chips. i've held since the 30s and it's been a boring compounder. the only question now is whether it keeps running, and that hinges on US rates into year end. holding, not adding at ATH.
$Oracle(ORCL.US) just had its worst month since 1990, down 35% in June on data center cost worries. it goes ex-dividend July 6 so a little more drift lower coming from that too. i like the OCI growth story but a falling knife with a capex overhang isn't my game. on the watchlist, not the buy list yet.
Michael Burry shorting $NVIDIA(NVDA.US) near 198 is such a 2026 headline. the man has called ten of the last two bubbles. stock's already 23% off the ATH. not adding up here, but i'm definitely not selling my core over a 13F filing. holding 🫡
Nvidia trades around $211 going into Wednesday's annual shareholder meeting, which lands early Thursday Singapore time. The stock has chopped in a roughly $208.6 to $214 band, and that compression tel...
while everyone argues about intel and apple, tsmc just keeps printing 30% revenue growth and runs the most advanced fabs on earth. the boring giant that actually makes everyones chips 💪
TSM +3% on monster May sales, AI demand extremely robust per management. this is the steadiest 1T plus name in the whole AI trade, screenshot this 🏹
still the most watched name in my feed even on a red day. NVDA holding while MRVL and INTC crater tells you where the quality is
bought the top last week like a genius and now this. this is fine 🐶🔥
+7.54% in a day is a lot to chase. love the custom silicon story but i want a pullback, not a vertical green candle, to add 🫳
happy holder but not adding AMD at 547 into a market melt up. high beta NVDA, i let it ride 🫡
not chasing up here, valuation is rich and a lot of good news is in. happy holder, not adding at the highs 🥱
not chasing here, valuation is full and the easy money in this cycle is behind us. happy to wait for a real pullback 🥱
this is not all or nothing lah. the giants with distribution and data survive, the weak small caps get eaten. winners and losers, not extinction 👀
big run already though. a single Jensen quote sending it this far makes me cautious, not bold. great company, fully priced for now 🫳
AMD riding the semis bounce and the TSMC pricing read-through. if AI silicon is tight, AMD sells everything it makes 💪
added a little into the fear this week, feeling fine about it today. not financial advice 🫡