@pmje73@tseides Best L/S podcast I’ve listened to in a while.
re the SM vs MM discussion, I’m curious, @pmje73, which product do you think better prepares analysts to be PMs themselves? Pros/cons of getting some portfolio responsibility earlier?
Skeptical whether labs go vertical fast enough to stake their claims in individual tasks. Feel like specialized model-agnostic will be able to build meaningful user bases before labs have appropriate focus
Stanley Druckenmiller:
"We're looking for AI use cases as opposed to infrastructure"
"Maybe the most interesting use case is a lab..."
"I don't personally believe they're gonna be in the token business for much longer...you can start selling tasks"
Best L/S podcast I’ve listened to in a while.
re the SM vs MM discussion, I’m curious, @pmje73, which product do you think better prepares analysts to be PMs themselves?
I had a fun catch up with my friend @tseides.
In general, I write more concisely than I speak. That’s why I like X.
I will probably add a few addendums here to expand on a couple of points.
First one -
I think big MMs being like big media Conglomerates now is a useful framing. Multi manager PMs are like TV Show runners for theee conglomerates. They need season upon seasons of shows. They need a team of writers writing shows and making the whole show impactful or they will get canceled. They cant rely on one good show in a season or one good idea in a show or one good episode or joke in an episode. To get renewed season after season you need to consistently write great shows all season long. MMs are like big networks, they have a lot of “airtime” to fill so they need a lot of show runners and a lot of different types of shows. If all the shows are the same, ratings for the network in its entirety will be bad. Sometimes shows get cancelled. Some networks are owned by conglomerates. Those conglomerates can further diversify by investing in different networks with different brands and strategies. HBO doesn't produce the same types of shows as ABC.
https://t.co/L4Ld9OQmHJ
SA incentives so misaligned in a down cycle. Semis bear market would mean subs get cut in half as generalists flood out and startup operators shut down. Have to assume house view is their bull case.
Semiconductor research firm SemiAnalysis has sharply LOWERED its global AI accelerator capital expenditure forecasts, with cuts of up to 41%:
2027: $770B → $627B (-19%)
2028: $1.054T → $758B (-28%)
2029: $1.352T → $796B (-41%)
2030: $1.531T → $923B (-40%)
That's a $608 billion reduction in its 2030 forecast alone.
SemiAnalysis also cut its cumulative 2026-2030 accelerator silicon shipment area forecast by 17%, with Nvidia accounting for 66% of that reduction.
Its 2028 HBM shipment forecast was lowered from 9.0 to 6.4 exabytes, a 30% reduction, while projected 2030 AI accelerator power demand fell from 122.8 GW to 80.7 GW, down 34%.
The firm cited slowing revenue growth at frontier AI labs, tighter capital markets, and lower assumptions for future chip and memory pricing.
Importantly, even after these cuts, SemiAnalysis still expects annual AI accelerator spending to more than double from $392 billion in 2026 to $923 billion by 2030.
AI semis/infra trade is uninvestable right now in my opinion, even 50%+ off highs (i.e. not meaningfully capitalizing ‘27/28 eps)…
> sure, maybe your numbers haven’t moved much yet and you can theoretically frame attractive upside on multiples just ripping back to highs
> but it’s 10x harder this time around, because now the burden of proof is on bulls to prove that accelerating capex is sustainable several years out
> that 20-30x your ‘28 eps only worked because everything, and I mean EVERYTHING, sounded perfect
> comes as now some of the AI trade’s biggest supporters are getting shaky (Jensen dethroned, Trump admin regulatory concerns, even SemiAnalysis capex revisions haha)
> if I’m an operator who (1) has capital markets getting more expensive, (2) no longer has Jensen calling me, (3) has elections telling me to slow down, and (3) my research providers predicting I WILL slow down vs. prior plans… whether I intend to or not, this will probably end with me actually getting more cautious
> cautious investment eventually causes the ROI side (better models, new use cases, etc) to slow too and then the downward spiral continues
> for public market investors, means to realize upside anytime soon you might need to hit a 10-leg parlay of positive data pts (bipartisan political support + successful IPOs + rates lower + new use case traction + reaccelerating DC buildouts, etc etc) in the window of time before ROI signals slow
Hard to see that upside case playing out soon. All ears if anyone has a rebuttal
#aitrade #aistocks $NVDA
nvda bull case:
-low pe on numbers that are too high
bear case:
-numbers too high
-export controls gonna zero China rev through SE Asia
-Jensen joins competitor or does startup and gets unlimited VC funding + selling nvda shares
-discounting to maintain share/margins down
-TPUs
Still think ending war is Trump’s best chance to pump equities over NTM, and therefore his most likely objective
Not my domain but any reason aero isn’t the group to buy? $GE $RR.L $ATI $HEI
AI momo getting harder to save after midterms and now Jensen debacle
BREAKING: Trump abandons talks with Iran, declaring the US no longer needs a deal with Tehran.
"We don't need a deal with Iran. Those days have long since passed. We've got their money, hundreds of billions of dollars, and we've got their Strait," he says.
Iran told Qatar in the past 24 hours that it would not negotiate over its nuclear program at all under any circumstances.
Qatar then relayed the message to the US, prompting Trump to say he no longer needs a deal, senior officials involved in the negotiations say.
Trump also told Vance to now "focus his complete and total attention on the massive corruption."
@JerryCap 90% of the defenses are completely disingenuous lol. Ppl whose paper net worths will live or die by the AI trade. Gotta get their exit liquidity somewhere
"Backstopped by NVDA" is really just "backstopped by Jensen" ... credibility now impaired for the whole complex. Ian Buck seems reasonable, but think it's hard to be buried below the king for decades and then suddenly command the throne. Time will tell
"Backstopped by NVDA" is really just "backstopped by Jensen" ... credibility now impaired for the whole complex. Ian Buck seems reasonable, but think it's hard to be buried below the king for decades and then suddenly command the throne. Time will tell
No true floor for $NVDA with Jensen out. Jensen sold vision better than anyone and will take long time for perm CEO to rebuild that trust. Off board too so he's going to have an incredibly well-capitalized *external* 2nd act. HFs not ripping off this short anytime soon