Value & Growth | AI & Tech stocks | Graham & Doddsville
Stock are my passion | 25+m
Entertainment purposes only | Do your own research | No personal advice
A deadly combo.
Diagnosis bias: the moment we label a person or a situation, we put on blinders to all evidence that contradicts our diagnosis.
Confirmation bias: we only accept new information that agrees with our diagnosis.
US government: Claude is too good. Must be a security risk. Stop non US citizens from using it.
China: Kimi is even better. No problem if it is open source.
Similar to the panic over DeepSeek R1, some uneducated people think Kimi K3’s use of linear attention (KDA) is bad for NVIDIA, HBM, DRAM, and networking because it has relatively lower KV-cache requirements. The opposite is true, and we explain why below. 👇️ 1/8🧵
1/ ASML just beat and raised guidance. That's the headline everyone will run.
But the number that actually matters was sitting in the deck (not the press release):
Taiwan went from 23% to 30% of ASML's machine sales in a single quarter. (hint: TSMC prints tomorrow)
The beat wasn't more machines. It was service.
Net system sales (~€6.6B) came in roughly in line. The ~€300M beat over Street was almost all Installed Base Management: service and upgrades, which hit €2.8B.
IMO this means the fabs are running their existing EUV tools hot. That's a direct fingerprint of AI fab utilization, and it's the stickiest, highest-margin line ASML has. It's why margin printed 54% vs 51-52% guided.
$IBM is getting smoked pre-market after releasing preliminary numbers... stock down -17%
Looks like $IBM is pulling down other software stocks (ie $NOW $ADBE $CRM $INTU $ACN $ESTC $SNOW etc) because of the following commentary...
"In the last few weeks of June, we saw clients shift their quarterly capex spend toward servers, storage, and memory purchases to secure supply-constrained infrastructure ahead of expected price increases. This dynamic impacted client buying patterns. While we anticipated some supply chain related impact in our expectations, we did not anticipate the magnitude of the capex reprioritization."
Bad commentary for software names... good commentary for memory, chips, and other hardware names.
@SlootLau Is 'gelegitimeerde diefstal' geen oxymoron? Vermakelijk boek: Taxtopia; aardig inzicht: u betaalt eigenlijk alleen belastingen om de inflatie binnen de perken te houden (bij landen die controle hebben over hun eigen monetaire systeem).
$AMAT Applied Materials: holding this stock feels like holding $AMZN through the internet bust: the business is firing on all cylinders, but the sentiment is shifting (temporarily, I think). I'm going to hold just wondering when to use my pretty big chunk of dry powder burning in my pocket.
@EhrmantrautCap_ Hybrid bonding will eventually get implemented more broadly but the question is whether it's fast enough for competition not to catch up; otherwise I think already too much optimism might have been priced in. $BESI
AI startup Perplexity is planning to use Nvidia’s new standalone Vera CPUs, with the company saying it found that Vera carried out agenting coding tasks 1.5X faster than traditional CPUs.
$NVDA $AMD
@NicoInberg Burry’s warning is valid for cyclicals, but I don’t think it’s peak earnings yet! Demand is outpacing capacity growth for at least the next 2 years. DYOR, cycles can turn fast.
@StockSavvyShay@VanquishTrader Data, talent, compute: Alphabet? I own both ($META and $GOOG); I'm not sure about Zuckerberg (and his values). $META is a small position for me.
Uniqlo, the Japanese retailer that has built its business on everyday essentials, is embarking on its next phase of international expansion. Harry Dempsey explains how its strategy could reshape the global apparel industry. https://t.co/f3uyQHLSQw
$AMAT remains DB's top pick, benefiting from high-volume DRAM and foundry investments. PT raised $550 to $680.
$LRCX also remains a Buy, with strong leverage to the multi-year fab expansion cycle. PT raised $325 to $385.