Realist retweeted
Equities: Surprisingly little fear
A pitch-black week lies behind us. Our concern about the surprisingly good sentiment was justified, but the extent of the price slump and its relentlessness caught even us off guard. All the more astonishing is investors’ reaction to the plunge: no panic, no panic selling and consistently positive medium-term expectations. Can this end well?
Read the full report @ sentix.de
Get free access by active survey participation.
Realist retweeted
Over the past three quarters, Nvidia deployed 58% of operating cash flow into equity investments and M&A, while off-balance-sheet long-term debt and lease liabilities rose from $214 billion to $422 billion (and $530 billion including land, power, and shell guarantees) from Q1 to Q2 2026. cdn.jpmorganfunds.com/conten…
Realist retweeted
GS: Despite the hedge fund unwind, broad equity investor positioning remains elevated.
Realist retweeted
US y/y percentage change in real disposable personal income... the red is basically the wealth effect... how long can that last as the S&P 500 holds in correction territory? EY
Realist retweeted
Anthropic sort of makes money.
OpenAI is in financial distress.
Thinking Machines and SSI dont make money.
A long tail of "neolabs" also dont make money.
What do you think happens when global training demand from VC-clown backed startups implodes because Jensen backstop gone?
Realist retweeted
Inflation expectations are falling fast.
Via @MikeZaccardi
Realist retweeted
Don't underestimate the size of the EPS revision narrative we are about to see... S&P Op Margins HATE falling pricing power.
I love that this is basically “Goldman Sachs spent a year trying to solve one of tech’s biggest headaches: how to commit legal fraud”
Goldman spent more than a year trying to solve one of tech’s biggest headaches: how to pay for AI chips without wrecking corporate balance sheets bloomberg.com/news/articles/…
Realist retweeted
Semiconductor earnings revisions breadth is rolling over. Strange - because surely they are not cyclical any more in this new era of infinite AI?
zerohedge.com/the-market-ear…
Realist retweeted
Denmark’s PM Mette Frederiksen:
There is an enemy growing within our own society.
These are supporters of Hamas. They are people who support legislation that places the Quran above our democracy.
Realist retweeted
I'm going to help all of the AI bulls out:
The most likely reason for the compute glut and massive malinvestment is that the people making these investments made a gamble thinking it would work out like it often has. The tech CEO's couldn't come out and tell investors they are going to make the most expensive gamble of all time and it might be a 0. They had to feign CONFIDENCE.
And for a hot second after the nexus of Claude code, agents, and the tokenmaxxing psychosis, it looked like the gamble might just pay off.
But now 6 months later, we have more data. There is no moat. Moore's law and compute efficiency gains are completely overwhelming Jevon's. Revenue growth curves are flattening fast and we don't have the next use case to absorb the trillions of dollar of compute supply coming.
With such massive investment looking more and more like a 0, and capex commitments so large that it actually might threaten the solvency of some of the worlds largest companies, the tech leaders couldn't just come out and say "we misjudged". They will look so foolish and that is the least of their worries. The size of the commitments they made are now FORCING them to keep up the charade to finance their commitments and stay solvent.
What once was possibly a good faith gamble on exciting new technology is now an existential fight for survival. We are at the stage where nothing a tech CEO says can be taken at face value.
They will tell any lie needed to prop up the bubble a bit longer, save their company, their share price, and their legacy while they pray for an AGI miracle. They will collude with each other to present a unified narrative and create an illusion of compute scarcity. They will eventually cook the books (if they aren't already) rather than be the person that bankrupted one of the greatest companies of all time.
If there is one thing I can implore the AI bulls to rethink - it is that these tech leaders are telling the truth and sharing all of the relevant information. Think about their incentives at this stage and stop hitting the replies with the next iteration of "BUT JENSEN SAID". We now have ample evidence Jensen is a liar.
Don't be a moron and replace "Jensen" with some other tech liar. And stop saying, "you think you know more than "insert tech CEO". I have far less inside knowledge about their industry and business but can bet I have a lot more expertise in capex cycles and incentive structures.
Tech CEOs know far better than I what is going on in their businesses, but to think that YOU do because of what they say publicly is comically naive. They have NEVER been reliable narrators, and now they have an existential crisis requiring them to tell whatever lie is needed to stave off implosion while they quietly dump their bags on you.
Mark Zuckerberg says Meta built AI compute before it knew there would be demand, and the bet only fails if AI overall fails:
"Some of the more complicated questions that we had were like, all right, so we're training this model. And we know we want to train the model, because this is key. But how much product demand do we think we're going to have? So now it's like, all right, well, the product doesn't exist."
"There was a moment where I think a bunch of people were like, yeah, maybe we should just be a little more conservative. And it's like, no, we're doing this."
"I don't actually think it's quite as risky as this. I mean, the numbers are big."
"If AI works overall, plenty of other people want all our compute, right? So they're willing to pay... if the cost is whatever, $15 billion a gigawatt, we could get a lot more for that."
"In order for it not to work out, it's not just that we have to fail. AI overall has to fail."
"We still have all these other uses to productively use compute. So in the grand scheme of things, we were definitely building out this compute for inference ahead of knowing that the product was going to be ready."
- Mark Zuckerberg, Meta founder and CEO, with Alexandr Wang (@alexandr_wang) on Cheeky Pint, uploaded October 12.
Realist retweeted
Part 1: consumer staples, bonds, high yield and now Dec 2027 sofr futures, last 20-30 days, all screaming growth shock is near. Markets are speaking, it’s time to listen.
Realist retweeted
FEDWATCH: “.. S&P 500 ended the week 1.2 standard deviations below its 200-day moving average
.. However, history suggests caution: .. in 2015, 2018, 2020, 2022 and 2025, the ultimate low came only after both the S&P 500 and Nasdaq fell to within about 3 standard deviations ..”
Realist retweeted
There's an old saying that nothing good happens below the 200-day MA
17 of the 20 largest drops ever for the S&P 500 started below the 200-day
Realist retweeted
And to think, all this could have been avoided if Jensen watched the 20 minute HR compliance training video
Realist retweeted
You know what capex cycles are? A leading indicator for future impairment charges. But the market is already well ahead of it.
Dust off those accounting rules on carrying value vs. book value for datacenters.


































