Flash Philo on Memory
The cure for high prices is high prices. June 2026.
"The first is the extreme brevity of the financial memory. In consequence, financial disaster is quickly forgotten."
–John K. Galbraith, A Short History of Financial Euphoria
We published on memory a month ago 👇
Philoinvestor
This is how I concluded the piece.
Memory manufacturers are perpetually doomed to suffer from high capex and cyclicality. For now, they are raking it in big – and spending a good chunk of that money to expand their production capacity..
American AI is still in maximalist mode, trying to brute force everything with more compute – and that benefits picks-and-shovels companies
When the Gestalt Switch of AI shifts to minimalism and elegance – everything could be turned upside down.
I have some ideas how and when that could happen – but I will save them for an upcoming piece.
Micron Technology: MU reported earnings last week, and the world went mad!
Blowout sales, earnings and guidance. This is all memory bulls needed to run to the streets and proclaim their intellectual and P&L superiority!
In other words, AIpoors need to understand that they are stupid, that they don't get it and that things will stay as they are.
That's the narrative doing the rounds as of late.
You can't even raise doubts about this AI cycle else you will be attacked and labelled as stupid, or worse 😂
And this is why many financial commentators just shy away from talking about it – and when they do, they water down their words just enough to not offend any of their followers and subscribers. Or more importantly, get caught on the wrong side of their views and lose aura points.
The thing is, this is not what I am here for. I am naturally undiplomatic and anti-political – I call it exactly as I see it.
But besides that, if I was diplomatic and hedged whatever word I said or wrote – you would not get any real value from me.
Case in Point: If I watered down my words, this would never have been published 👇
Philoinvestor
- BTC down 40%
- MSTR down 80%
Now let’s examine Micron’s results before placing them in the broader context of the AI cycle.
The Actual Narrative
I love the internet because you can discover the narratives and viewpoints of people that you probably would never be exposed to, without the digital world.
Mr. Casrow below is one of the many Memory Bulls doing the rounds on X and celebrating his wins.
The narrative from the perpetual-memory-bull side seems to be that we are in a period of infinite demand for compute and intelligence – and therefore memory.
"Only a handful of companies produce memory at this quality and therefore the rest have no choice but to pay any price to the memory producers."
"YOU LITERALLY HAVE NO CHOICE!"
I note that in this cycle, the dominant genre of market participant is that of the technologist. Let's call it the poster genre of the cycle.
The majority of them made so much money being long Semis, Palantir and more recently Memory that you just can't touch them. They are the people bidding up Real Estate and making your favourite hotels simply unaffordable.
They made the most money, they are obviously the most successful in this cycle and their story is the one being told the most – and that's normal.
People like Leopold Aschenbrenner are being deified and used as an example for how to play this market right now. The hype is real.
But what these technologists do not want to understand is that paradigms shift and cycles change – but that the human mind is much slower in adjusting.
Palantir 🔻
Let's take the example of Palantir.
Before Hyperscalers and Anthropic ruined the market for memory – giving rise to yet another bottleneck trade... Palantir was the hottest thing in AI, up like 20X or 40X depending on where you bought.
The company beat expectations quarter after quarter, and commanded a sky high valuation that took forever to actually weigh on its stock.
So someone who made 20 or 50 million unrealised on Palantir is now down ~50%.
"If I made 50mln dollars on Palantir I don't care if I am down 50%" you shout!
Why does this matter though?
AI Winter 🥶
Many names in the AI complex are taking a lot of pain – and so are the investment portfolios of those holding them. We literally have 20% drops from 10X ATH and holders are crying about drawdowns – the entitlement is crazy.
Big Tech/AI is still allocating everything they have (and more!) to the AI infrastructure trade.
But IMHO we're approaching the point where the first of them will officially announce cuts to the budget.
Hyperscalers will not stay put buying whatever they are fed from the picks and shovels players hoping for ROI. Competition in AI is fierce and ramping up even further!
If you believe that cost is the ultimate differentiator, that competition is real and that creative destruction is a force to be reckoned with – then you believe the sword of Damocles is hanging above the head of the AI and Compute Trade, like so.
Scratching the Narrative
Now let's see what's really happening in this business. We will touch on cyclicality or lack thereof – and pricing power between memory producers against those using memory in their products.
Who holds all the cards?
“We have never seen a component price increase this much, this quickly,” Apple said in a statement. “We have shielded our customers from these increases so far, but we have now reached a point where we need to begin raising prices on a number of products, including today’s increases for iPad and Mac.”
Notice how Apple did not raise prices for the iPhone. That's because they are scared to in case this boomerangs and they print a 30% drop in sales for the next quarter! For now, they are just testing the waters.
But also note how this was an opportunity for Apple to extract more from their customers and shift blame to someone else.
After Apple's thinly veiled blaming of memory manufacturers for this price increase, Micron hit back saying Apple was overly aggressive in previous cycles and underpaid for memory...
🥁🥁🥁🥁
And because of Apple's squeezing Micron, they could not build enough memory factories back then.
And there we have it, admission from Micron itself that Memory is not only cyclical, but highly cyclical.
Don't get it? Read on.
