Pinned Tweet
Cerebras ($CBRS) trades at ~10x 2027 sales, priced like the chipmaker that finally breaks Nvidia's grip on inference.
Its own 10-Q describes something closer to a neocloud.
And its CEO pleaded guilty to a felony accounting control violation in 2007.
Thread👇
Technology stocks are going to be uninvestable for at least the next 3 years, maybe even 5
The hardware side of the industry is going to have to overcome the biggest overbuild in its history. It will be a knife fight for market share and margins will be thin
Labs are going to see huge revenue declines as the cost of compute craters from oversupply. Compute will basically be free so all you are paying for a is software wrapper on top of that
The AI bear case on software was always real, it is just a slow moving transition. With massively abundant and cheap compute, software will be even more under threat of commoditization
Obviously there will be some individual winners, but we are exiting the era of technology leadership in the stock market
Remember $SOXL?
Did anyone have the balls to short it?
Hilarious chart
2
4
592
Public markets have just become a dumping ground for for VC and PE fake valuations.
It's too bad because they used to be a place a small company could legitimize itself and the public could get massive upside gains from finding good companies early
Now private capital just hoards the gains for themselves. It seems extremely unamerican (or very american depending on your stance)
Top IPOs this year all down massive: $ANT $SPCX $CBRS $INIO $MAIR $FRVO $QNT
$TLT is probably the best place to hide right now with everything so uncertain. The AI complex has been keeping markets tight and helping keep unemployment low. If the complex falls apart, inflation and rates probably come down fast
It has probably been one of the worst 5 years ever for this trade, which means it is a great time for it to start to outperform
4
432
I seriously wonder how much of today was a bunch of retail traders who need to be fully invested literally every minute of every day, dumping their stocks to invest in $ANT the minute it became tradeable
2
4
664
"The bank of AI is under supervision from federal authorities"
Amazing clarity in this short interview
I believe Nvidia is the biggest case of industrial espionage and fraud in the history of the US stock market.
The bulls still defend the AI trade by pointing to chip prices, because used GPUs have held their value on both the contract and spot markets.
I believe keeping those prices buoyant was one of Jensen Huang's goals all along.
There are tens, maybe hundreds of billions of dollars of chips sitting in warehouses collecting dust, and the only real buyers were loss-making, VC-funded startups.
Now the bank of AI has federal authorities in the building and its chief loan officer is gone, so the money that funded those buyers stops flowing and the chips have nowhere to go.
Tom Keene asked me on Bloomberg if anyone can measure the residual value of the silicon in these data centers the way you can price a used Boeing 737.
My answer: You can't. And it's about to get much worse, because those chips will be selling on eBay by the truckload.
3
536
Where do people think $QQQ ends the year at this point? It is currently up 5%
31%Higher than 5%
31%Between (5%) and 5%
23%Less than 5%
15%Just tell me what to do!
13 votes • Final results 1
186
I have a base rate calculator I built in Codex based on @mjmauboussin's "Bayes and Base Rates" paper
I put in $ANT's historical revenue growth and assumed they hit their target for the end of the year of 10x y/y growth in revenue
Here is what base rates project for ANT's revenue going forward:
-40% in 2027 (sell side expects 151%)
-20% in 2028 (sell side expects 88%)
-12% in 2029 (sell side expects 35%)
I asked if any company has ever achieved the level of growth projected by the sell side. This was its answer:
"No close historical match exists once ANT reaches scale."
AOL, Alphabet, Block, and Tesla were the closest to meeting it.
Do you think ANT defy the historical base rates?
2
746
No position, but seriously how is $PANW not the most overvalued stock in the market?
>$300b market cap for a company with $14b of revenue and $5b of EBITDA
Even looking at 2031 projections, the expectation is for $9b of EBITDA
I get that they can keep beating estimates, but it just seems like a horrible r/r from here
$PANW acquisitions get bigger every year
*PALO ALTO NETWORKS TO BUY APPLIED COMPUTE FOR $4.5B: WSJ
4
6
2,121
$CBRS down 7% today. what did Feldman say in those closed door meetings?
8
18
9,922
Anyone know why $NTSK is worth $9b?
Is it just the cybersecurity halo?
1
3
612
So a company that has arely been able to grow and where most divisions are in perpetual decline is going to grow top line 5-8% from 2028-2030? $MH
This makes absolutely no sense and this will never happen. Platinum Equity @PlatinumEquity is clearly about to hit the eject button on this turd that they own 86% of







