Pythia Cap: Partially Conductive

83.9K posts

Pythia Cap: Partially Conductive

Pythia Cap: Partially Conductive

@PythiaR

Multibillionaire space pirate. CIO of Vader Capital LP. Activist engagements include Jeddha and Alderaan.

Des Moines, Iowa Katılım Haziran 2019
1K Takip Edilen70.8K Takipçiler
Pythia Cap: Partially Conductive
I think that's right. I think @viggy_krishnan is right when he says that semis need the hyperscalers to work as a group before they can really OP again because the market needs more confidence in capex increases at the hypers and for them to do that, they need to show returns/growth acceleration that will be positive for those stocks. But I think over the next ~6-9 months that probably starts to happen? Also worth noting on memory: Hynix capital return is a % of FCF, which comes after capex, so technically semicap gets paid before shareholders ;)
English
0
0
1
24
Franny
Franny@Franny8·
@PythiaR I like point 2 a lot. I also think we have maybe 2 years left with a lot of highs and lows. This all changes if big tech cuts cap ex in mass but i don't see this happening outside of say MSFT
English
1
0
1
19
Pythia Cap: Partially Conductive
If the Koreans took Memory too far in one direction, why wouldn’t they take it too far in the other?
English
1
0
19
3K
Pythia Cap: Partially Conductive
I think the cycle will behave like every other memory cycle in history: the memory guys will end up spending 100% of through cycle free cash on capex, ~70% of which goes to the big tool companies. What's going to be different is the duration/magnitude of this cycle. Where I'm going to be wrong is if they show collective restraint on capex and return gobs of capital to shareholders and they might for a little bit (~1-2 years) but: 1) every commodity cycle people think everyone is going to be restrained and then they aren't so I don't buy it especially if China is involved because they have a history of wrecking the economics for everyone else in literally every industry they enter; 2) the LTA's incentivize volume growth because they limit price, so, I think at the very least it's going to be a strong few years for semicap; 3) I suspect the Koreans are going to return significant capital to employees, I'm not so sure about shareholders.
English
1
0
1
47
Franny
Franny@Franny8·
@PythiaR Why semicap over memory complex? R/R?
English
1
0
1
27
Pythia Cap: Partially Conductive
American investors bidding $SKHY to 40% premium to local while Korean investors send the local down every day because of cascading margin calls kind of tells you what you should be doing ;)
English
4
0
21
2.3K
Pythia Cap: Partially Conductive
Well, I think: 1) DDR is a commodity that will normalize. Can't have 90% margins on that (which are above HBM) forever. Demand destruction in mobile, PC + easier supply additions etc. My working thesis is China will eventually fuck everyone here, but it may take longer (e.g. CXMT & YMTC probably going to be producing for internal Chinese needs for a while yet, not threatening HBM). 2) HBM obvs. eating wafer capacity for #1 which is why that's tight, don't see HBM demand slowing down yet. Don't see the trade ratio getting better either (e.g. MU saying its' going up every gen). I'm less worried on that S/D picture here than DDR. 3) I don't think anything materially changed between June 30th and today, yet the stocks are down 40% in Korea. IMO it's all technical, I bet in part by people converting from the Korea listing to US ADR, but also Samsung pre was "weak" (even though they didn't say why they missed), Korea broker reports on Hynix missing (but Hynix roadshow was so bullish...). So I'm okay buying at Korea prices. I've been buying $DRAM which is mostly Korea, don't really want to own Micron but here we are. But that said I think it's also a tougher road ahead, it'll chop. I'd like to see capital returns, think that would help, but I'm very clear I own it cuz I think there's another 50-100% upside in these over ~18 months, I don't think they're going up 300% again. And sizing wise, I have 15% ASML, 5% ACMR to 5% DRAM (and ~2% RAM, the levered DRAM because I like the occasional degen bet) so it's very small, my main related bet is still semicap.
English
2
0
2
77
Franny
Franny@Franny8·
@PythiaR You getting more bulled up than me on memory lol.
English
1
0
0
67
Stocksailor
Stocksailor@stocksailer·
@PythiaR well we can't buy local shares.... obv locals shares are a better deal
English
2
0
0
233
Pythia Cap: Partially Conductive retweetledi
Quantіan
Quantіan@quantian1·
I have never, EVER seen language like this from a $FRMI board member in a resignation letter. Proactively assuming the company will lie to investors about why you resigned?
Quantіan tweet media
English
10
4
56
7.5K
Kenneth Chan
Kenneth Chan@iamkennethchan·
Vancouver needs the courage to let neighbourhoods evolve. The Villages Plan does just that. The objection to it no longer appears to be about the specific form of growth (high-rise density vs. low-rise density), but about growth itself. #vanpoli #vanre dailyhive.com/vancouver/vanc…
English
11
6
61
6.4K
Pythia Cap: Partially Conductive retweetledi
Aniruddh Mohan
Aniruddh Mohan@aniruddh_mohan·
LCOE as a metric has so many flaws, but the yoy change tells you where the market is headed Interesting trends here from Lazard’s 2026 update - New combined cycle gas LCOE at 15-year high - Wind & Solar also go the wrong way on costs, solar costs up 18% yoy - VC/academic viewpoint might be that wind/solar only gets cheaper with time but not a surprise to folks tracking PPA market, LevelTen says wind/solar PPAs at their highest ever since their data began in 2018 (!) - Solar continues to get more expensive due to tariffs, cost of capital increases - Storage also reverses years of declines, went up notably over 25-26, new FEOC concerns repricing supply chains - Lazard: "relative economics of existing generation have improved as rising new-build costs across all technologies, together with execution challenges tied to supply chains, inflation, tariffs, permitting and macroeconomic uncertainty, have made replacement capacity more expensive and difficult to deliver"
Aniruddh Mohan tweet mediaAniruddh Mohan tweet mediaAniruddh Mohan tweet mediaAniruddh Mohan tweet media
English
8
27
102
26.3K
Pythia Cap: Partially Conductive retweetledi
Ed Bradford
Ed Bradford@Fullcarry·
A negative core MoM
English
3
7
46
17.3K
Pythia Cap: Partially Conductive retweetledi
High Yield Harry
High Yield Harry@HighyieldHarry·
Do not fall for this guy's crocodile tears. He just needs your money again.
High Yield Harry tweet media
English
21
11
537
31.6K
Pythia Cap: Partially Conductive retweetledi
🌿 lithos
🌿 lithos@lithos_graphein·
Jim Cramer 20 hours ago: IBM has a "90% probability of a Q2 earnings beat"
🌿 lithos tweet media
English
9
6
69
28.5K
Pythia Cap: Partially Conductive retweetledi
Thijs
Thijs@cdngdev·
i gave 5.6 sol access to my camera roll and had it extract pictures of every piece of clothing i own from my photos then, told it to find new outfits for me and render them on me with gpt-image! its kinda cool to see your entire wardrobe in a collection like this
Sam Altman@sama

i'd love to see interesting things people have built with 5.6 sol. i will send the person who made the coolest thing a special gift from the openai archives.

English
332
569
15.3K
3.1M