The Healthy Correction Phase 😂
Everything is unwinding.
"This business is filled with big egos. Don't have a big ego, have a strong ego. It's not the same." –Philo

So, how are markets doing lately?...
Well, Nvidia is down 38% for the month, with one more trading week left. The Nasdaq closed down 10% for the week and the AI trade is imploding.
But I see many still in denial. Denial about the setup, about the actual process that brought us here, and denial about what a bubble really is.
But don't get lost in the details, as I keep on saying: operational success is what matters. You don't want to philosophise your portfolio to death.
Many will in this cycle, and many have already.
So it's time for another Breakout by Philo to put everything together in a short and sweet weekend post.
Why a Breakout by Philo now? 😇
Because I haven't seen charts this beautiful in a long, long time...
And while every market commentator comes out to convince us that this is just a healthy correction – markets and reality have a different opinion.

KOSPI 200
The KOSPI peaked in June, after Micron's blowout quarter and their long-term agreement innovation.
For more on that, read Flash Philo on Memory, and my tweet after their earnings here.

The index is down >50% peak to current value – with no major bounces as it hit support levels, besides the first one in July/August.
If you bought the narrative about memory no longer being cyclical, or that P/E ratios somehow matter in a linear fashion.. You got burned here.
What about Semis?

The VanEck Semiconductor ETF peaked at the same time – but was held up by Nvidia and AMD, until recently.

The weakest Semi name has been unsurprisingly Intel, with even long-term bulls throwing in the towel publicly about the name after recent performance.
With everything that's happening in AI: expect more news about its major capex program and its fab buildout..

AMD performing much more strongly, and still holding the level >$400, for now.
And of course, we left the best for last: NVIDIA, or rather, the Central Bank of Nvidia.
This is the term we used to describe the importance of Nvidia in the AI trade. In fact, we published three pieces under that name, the first one being in September of 2025.
That's why we used the term Central Bank of Nvidia in the first place.
Read the first part of the series here – now unlocked from paywall.

NVDA stock made new highs in October and then reversed and never looked back (up) again. The stock is down 50% peak to current price.
A month ago we published Part #3 of the Central Bank of Nvidia.
Reading that would have maybe saved you a few trillion dollars in your PA, tongue in cheek. Note: the conclusion to the piece hasn't fully played out yet, there's more to come.

The Neoclouds
In the Central Bank of Nvidia series we explained how integral Nvidia's involvement was (and is) for the Neoclouds.
And few AI sub-sectors have been more controversial than the Neocloud trade.

NBIS is down to $80 from $300 and CRWV is back under $40 from $180.

This severely inhibits their ability to raise more equity capital – in order to fund their massive capex programs and forward commitments. Clearly, their crashing stock prices affects their ability to raise debt capital too.
This is catastrophic for these companies
And puts them in a position where they need to 1) restructure and/or 2) find other strategic partners to help them adjust to this new reality. Maybe Michael Saylor will oblige? Tweet.
If you don't understand how this affects the whole AI trade in a domino like fashion, watch this whiteboard video I made last August.
Simple...


The Indices
The Nasdaq on the Weekly looks utterly broken. Sideways for months and then just breaking levels to the downside without any reprieve.

Even Google is crashing – having over-hyped its TPU angle, and with their Search segment now facing headwinds.
From >$400 to $265 in a few months has put a dent in most portfolios out there...

People are hurting out there: but look at the bright side. Maybe houses in San Francisco will get a bit cheaper!
Meanwhile the NIKKEI is also breaking levels to the downside, decisively. Charts look broken all over the place.

Conclusion
This is not business as usual.
Resist the urge to think linearly about this – and slapping P/E ratios to these names in order to rationalise your bias.
"NVDA is just 16X earnings and Micron is whateverX earnings..."
What value does P/E have when you don't know what the E is going to look like? Don't be stupid, please.
And don't connect AI having value as a technology to AI stocks necessarily benefitting from that. Higher-order thinking is the name of the game: Creative destruction and race to the bottom.
We've been talking and writing about this for literally years. We've published several pieces on all those dynamics and it's time for you to brush up on your reading.
Final Thoughts
Finally, what if the market decides the Hyperscalers should get more discounted from here. Mag7 has lightly sold off, but nothing compared to what can be.

Further Reading




