I guess the mkt bid $WMT to 50x EPS on egg price inflation
English
SouthernValue
6.7K posts
SouthernValue
@SouthernValue95
Infinitely curious. “Spend each day trying to be a little wiser than you were when you woke up.” - Charlie Munger
Boston, MA Katılım Haziran 2019
1.7K Takip Edilen15.6K Takipçiler
@kylekuzma @evrgn11112231 forgot to switch to your burner
English
Yep wait till their data center build out 🔥🔥🔥 heavy capex
Oguz Erkan@oguzerkan
$META trading at 18x forward earnings while growing +20% YoY… What’s the bear thesis here? Really?
English
@j_fishback What do you have against AI data centers as long as they solve for their own power and water?
English
BREAKING: I’m running to succeed Ron DeSantis as Florida’s next Republican Governor to make life more affordable for you and your family.
I’ll stop the H-1B scam, tell Blackstone they can't buy our homes, cancel AI Data Centers, and abolish property taxes.
English
@Larryjamieson_ C.C. Wei $TSM
Tricia Griffith $PGR
Tim Cook(Ed) $AAPL
Brian Lane $FIX
Jamie Dimon $JPM
Hok Tan $AVGO
Mark Leonard $CSU
Satya/Amy $MSFT (yes even still)
Larry Culp $DHR $GE
English
What’s the best manager/management team you’ve ever seen?
English
@SouthernValue95 DRAM is not made by TSMC / NVDA
What is this low level of understandings of the market lol
English
Spacex aimed for Mars and became a Neocloud. At least should be a very profitable one.
If Terafab aims for TSMC and NVDA can it at least help end the DRAM shortage?
English
Will be interesting to see which company Elon attempts to acquire first: Intel, Substrate, Cerebras, Space Forge or Etched
English
@JerryCap Yea, they will work when cyclical stuff driving index now (tech capex cycle) does not
English
@SouthernValue95 Don't forget rates going 0 to 5 and more semi value is next two years vs compounders in TV (more rate sensitive mathematically)
English
It is a very good year for earnings. $SPY EPS grew +28% yoy in 1Q and are expected to grow +22% for CY26.
Largely driven by Mag7 + their capex beneficiaries, but even ex-Mag7, earnings growth is in the high teens.
English
The index is growing earnings 28%… if you’re wondering why your compounder growing EPS 8-12% that was trading at 30x has de-rated… that’s why!
English
Earnings & revenue beat/miss, and yoy growth by sector. Led by tech & comms but broadly strong.
English
Gone for a few days and…
Did I resurrect / provoke @GavinSBaker who is now spitting fire and roasting people in comments all over the place on here lmao 🔥🔥
lol at people freaking out about (or downplaying) $META ’s $RDDT knock off… this is just the beginning (infinite apps thesis inning one)… for those saying “threads was a failure” don’t look now but it’s not? #5 most downloaded app in app store today. The Forum app from Meta is cool but the best part is the “ask” tab which is an ai chatbot that specifically only pulls from the Groups community and can be / likely will be a very interesting standalone chatbot app (AI answers but only summarizing and linking answers given by real people over decades). the Forum app should default to grant you the option for posting under an avatar though vs your real name if the goal is to drive a real engagement step change (same thing should be upfront in threads).
re Meta layoffs - the handwringing and entitlement by people on here in the tech ecosystem about this is quite insane… very out of touch group. separately but related - under discussed yet but very interesting phenomenon is Meta’s decision to move 7k engineers to exclusively focus now on AI products and data labeling. Not only are we going to have a workforce greater than the size of anthropic and OAI exclusively focused on building capabilities at scale that no one else really can, but will also get a lot of cheap / forced attrition from toxic employees who think they are too good for this kind of work.
Google i/o was super lame and confirmed my views on their inability to ship and that the race to win in consumer / agentic AI has been won (by meta) before it even started
I bought a software stock this week. anyone who has followed me on here or known me irl should know how big of a deal that is as i’ve long hated the space broadly. put me officially in the software bull camp now though as my exposure is now larger than my semi exposure.
any guess what i bought? will tell in time have work to do first.
what else did i miss?
English
@RHouseResearch COGS = maintenance capex. Think though the idea of needing $5T capex to support $4T of revenue… doesn’t work
English
Some crude AI TAM analysis:
- It was recently reported that $CRM is on track to spend ~$300M on Anthropic this year. That works out to ~$300/mo per employee
- A GitHub copilot customer I spoke to was quoted a 10x price increase in June ($300/mo)
- I myself happen to spend ~$300/mo in AI subscriptions
- Let’s use $300/mo as a proxy for decent penetration of AI into a knowledge workforce.
$MSFT 365 has ~450M users, $GOOG Workspace has a similar number (est., not disclosed). That’s a decent proxy for global knowledge workers ex-China at ~900M.
If every knowledge worker consumed $300/mo of tokens, that’s a $270B monthly or $3.24T annually. Would equate to ~5-6% of global white collar labor spend ($50T).
On consumer side: if every GOOG/META DAU consumed ~$20/mo of tokens, it would be ~3B * $20/mo * 12 = ~$720B. GOOG/META could basically fully subsidize this with their existing business models within a few years. Let’s assume $20/mo of tokens is covered some mix of ads, subscriptions, commerce share, and subsidization.
Combining the TAMs we get to ~$4T of token spend. Keeping to simple numbers, at ~50% gross margins, it would sustainably require ~$2T of infrastructure spend to support these tokens. That would be ~2x the 2026 rate of ~$1T of AI infra spend ex-China.
So from there:
- Is enterprise token spend too high or too low? Could argue token consumption will only grow for CRM, but also that cost per token will fall, and the avg company won’t consume that much.
- Is consumer token consumption too high or too low? Maybe consumers will be more willing to spend on subscriptions for this than in past consumer tech cycles. Maybe it’s too high if ai models don’t monetize.
But at least $300/mo per user for enterprise and $20/mo for consumer are tangible numbers I can get my head around. That feels like the bogey for infra cycle to have another double from here and not be a huge bubble, and probably need another double at least for the stocks to work.
Very crude round numbers throughout here. Appreciate thoughts and pushback.
English
@jdonovan42 @malinvested Depends on the models that capture the inference. So far most share continues to stay at the frontier, and those models are too large for local inference.
English
@malinvested @SouthernValue95 You think the inference infra will be mainly customer premise/edge or cloud? Or hybrid? I suspect edge for physical A.I. like plants, but not certain on rest.
English
Btw, we can debate the size of the token TAM, but eventually it has to be larger than the infrastructure spend to support it. If you aggregate consensus views for the major token spenders (clouds, $META, frontier labs), you don’t get anywhere close to $3-4T of incremental AI revenue in the next few years, even with the explosive growth at anthropic and the acceleration at every hyperscaler.
This means that either: 1) token “spender” revenue estimates are too low, or 2) infrastructure spend needs to peak sometime in the next few years to allow spenders to grow into their investments.
Note the call from @coatuemgmt last week that at some point we ought to see a shift in equity value capture from sellers of tokens to buyers of tokens. x.com/mollysoshea/st…
SouthernValue@SouthernValue95
Some crude AI TAM analysis: - It was recently reported that $CRM is on track to spend ~$300M on Anthropic this year. That works out to ~$300/mo per employee - A GitHub copilot customer I spoke to was quoted a 10x price increase in June ($300/mo) - I myself happen to spend ~$300/mo in AI subscriptions - Let’s use $300/mo as a proxy for decent penetration of AI into a knowledge workforce. $MSFT 365 has ~450M users, $GOOG Workspace has a similar number (est., not disclosed). That’s a decent proxy for global knowledge workers ex-China at ~900M. If every knowledge worker consumed $300/mo of tokens, that’s a $270B monthly or $3.24T annually. Would equate to ~5-6% of global white collar labor spend ($50T). On consumer side: if every GOOG/META DAU consumed ~$20/mo of tokens, it would be ~3B * $20/mo * 12 = ~$720B. GOOG/META could basically fully subsidize this with their existing business models within a few years. Let’s assume $20/mo of tokens is covered some mix of ads, subscriptions, commerce share, and subsidization. Combining the TAMs we get to ~$4T of token spend. Keeping to simple numbers, at ~50% gross margins, it would sustainably require ~$2T of infrastructure spend to support these tokens. That would be ~2x the 2026 rate of ~$1T of AI infra spend ex-China. So from there: - Is enterprise token spend too high or too low? Could argue token consumption will only grow for CRM, but also that cost per token will fall, and the avg company won’t consume that much. - Is consumer token consumption too high or too low? Maybe consumers will be more willing to spend on subscriptions for this than in past consumer tech cycles. Maybe it’s too high if ai models don’t monetize. But at least $300/mo per user for enterprise and $20/mo for consumer are tangible numbers I can get my head around. That feels like the bogey for infra cycle to have another double from here and not be a huge bubble, and probably need another double at least for the stocks to work. Very crude round numbers throughout here. Appreciate thoughts and pushback.
English
@BluthCapital Depends if this end up more utility-like, as internet and telecoms are, or a nice-to-have.
English
@SouthernValue95 CRM is facing a potentially existential product disruption; it *better* be spending more on AI than almost anyone. Imagine it loses a big account bc its AI chops weren't up to snuff and DynamicsCRM leveled up/rebranded and started winning.
English
