---
title: 'The Memory AI Stock Collapse is STARTING'
source: 'https://youtube.com/watch?v=8rDo0AjV6zI'
video_id: '8rDo0AjV6zI'
date: 2026-07-01
duration_sec: 1192
---

# The Memory AI Stock Collapse is STARTING

> Source: [The Memory AI Stock Collapse is STARTING](https://youtube.com/watch?v=8rDo0AjV6zI)

## Summary



## Transcript

How much should you be worried about the
memory stock collapse that could
potentially destroy the entire hardware
stock market? And given that hardware
stocks have really contributed to over
75% of the S&P 500's gain since April,
makes you wonder how strong can indices
hold up if we potentially see a collapse
in memory stock prices. Is that even
something that's likely to happen? Today
is one of the days that has people
scratching their heads given that
SanDisk and Micron are down over 6%
today. Dragging the triple Q's the
NASDAQ 100 off that 17 uh 715 level even
though we've got that SpaceX inclusion
coming up on July 7th. We got to break
down exactly what's going on here and
what you need to know about how
important these memory chip stocks are
and what their history generally tells
us. So, first things first, I meet Kevin
about a week and a half ago. I warned
that no matter what happened with Micron
for earnings, whether it was good or
bad, it was not something I was
interested in playing calls on. I made
that call not to buy calls. And now
Micron is down 17% from peak despite the
fact that they absolutely dominated
earnings. So, why is Micron down 17% and
falling as much as it is today when the
company's earnings are doing so freaking
well? Well, the answer has to do with
history. And no, not my history getting
a nod from the Pope, which was really
cool in Barcelona just about a couple
weeks ago. And yes, we do have a coupon
code to the Meet Kevin programs on
building your wealth. Of course,
remember live streams, trade alerts, and
the morning alpha report. Every morning
the market is open and expiring
tomorrow. But what matters is the
history of memory. See, when we actually
look at memory prices, people have a lot
of prognostications on what memory
prices are going to do. For example,
here's a guy called the money guy. He
runs a blog and he argued that we might
end up seeing a dramatic collapse in
memory prices in 2025. Fingers crossed,
hopefully not. Because he basically drew
this line going all the way back to
2005. drew a trend line of monthly
memory revenue and suggested that hey we
go through these ups and downs along
this line and if we get a little up here
we might get a down again in 2025. In
fact literally wrote over here revenues
seem to have followed the pattern of the
past and this could mean we'll have
another boom and bust cycle with a
sizable downturn in 2025. Let's hope
that doesn't happen again. And this was
written back when memory prices were
collapsing in October of 2023.
What's fascinating about this is October
of 2023 actually led to losses in the
memory divisions of Samsung and of
course the entire companies of SKHX and
Micron. In fact, Samsung lost 11.6
billion as recently as 2023. That's just
3 years ago. So, three annual reports
ago, we could say two annual reports ago
since we don't have an annual report out
for this year yet, Samsung lost 11.5
billion on their memory sector, Micron
as a company as a whole lost $6 billion
and SKHEX lost $7 billion
in a year. In fact, it got so bad at
Micron that their number one risk
factor, the very first risk factor that
they list on their annual report, which
is volatility in annual uh in average
selling prices is exactly what ended up
tanking their revenue in 2023. Take a
look at this. In 2023, Micron made about
$15.5 billion of revenue, but it
actually cost them $16.9 billion to
manufacture those chips. That means they
were actively running their chip
fabrications at a loss. chips require
basically high volumes and almost 100%
utilizations in the factory uh in their
factories memory chips to actually
sustain profitability and when you can
go from hey look we made $10.4 billion
during COVID you know 2021 over here
right after co I suppose and 2022 we
made $14 billion oh just kidding now we
lost $1.4 $4 billion in gross margin.
And actually, instead of making uh money
over here at net income, yeah, we lost
about $6 billion. So, every penny we net
in 2021, we turned around and lost in
2023.
This has nothing to do with Micron being
a good or bad business. It has
everything to do with risk factor number
one, volatility in price. And this is
why memory chips and memory chip stocks
really, in my opinion, are kind of like
oil and oil companies. Oil companies
make a lot of money when oil is
expensive, and they lose lots of money
when oil gets cheap. That's why it's
their number one risk factor. Listen to
this. Average selling prices for DRAM
declined in the high 40% range and NAN
declined in the low 50% range for 2023.
And since 2017, annual price changes in
DRAM selling prices have ranged from
plus 35 to minus the high 40s percents.
So in other words, they're basically
saying, hey, like over the past decade,
we have seen memory prices go up 40%
year-over-year to well 35%. To down
almost 50% year-over-year. down 50% is
bad because you know we could go up 200%
in prices but down 50% you know those
numbers they add up really quickly and
you stay in these holes. In fact they
indicated that since 2017 they say it
right here annual price changes in NAND
average selling prices have ranged from
nearly flat to minus 50%. So DRAM at
least had plus NAND was flat to negative
50%. average selling prices for our
products that declined faster than our
costs have recently had an adverse
effect on our business and operations.
So in other words, memory was in a poopy
dupy situation in 2023. And that was one
of the reasons a lot of these memory
chip stocks absolutely plummeted in
value and they sold for really really
cheap prices right at the beginning of
artificial intelligence. And then of
course in 2025 everybody got really
excited about a gentic AI which has
substantially more memory demands. And
of course I'm oversimplifying here but
what happens the commodity that is
memory which is exactly pinforpin
replaceable with a different company's
memory chips. It is a commodity uh well
skyrocketed because now we have
enthusiasm over aentic. But we have to
take a look at the last two quarterly
reports to see what kind of pressure we
actually have on Micron and the
importance that they keep delivering to
keep this market propped up. Now, quick
reminder, yes, we are expiring coupon
code tomorrow. It's coupon code pope.
You could join the alpha report every
morning where we talk short-term,
medium-term, long-term trade alerts.
Part of that is because over at Reinvest
AI, so reinvest.co, we are releasing the
valuation product tomorrow. This has
been in beta since about Thanksgiving
with more or less likely deals, but
we're actually coming out with where we
think deals can rank based on how much
money you can make off of them. So,
we're really excited about that. Both of
these products will have an expiring
coupon code tomorrow. So, the Meke Kevin
membership as well as the Reinvest AI.
That's the house hack uh product over at
reinvest.co or meet Kevin.com for the
Meke Kevin membership. So, what do we
need to know when it comes to Micron's
last earnings? Well, take a look at this
right here. You've got Q3 Micron. All
right, these are the numbers that just
came out. What I want you to do is just
see the difference between these two
quarterly reports. I flash them back and
forth. I basically perfectly aligned
them for you. This one right here with
the lack of highlighting. I'll use some
red over here. This was last quarter's
report and it was really good. In fact,
if as a course member you go over to the
stock tab, you can see I highlighted the
crap out of these documents. Just to
give you some ideas over here, this
company put 11.7 billion into capex in
just Q1. They expect to spend $25
billion. They're expanding their high
bandwidth me uh memory facility in
Singapore in 2027. We expect more
facilities to come online in uh what do
we got over here? We've got the Idaho uh
production um or or Idaho memory
facility is expecting to come online
midcal 27. That would be next year. Then
we've got another fab coming out at the
end of 28. We got another fab coming out
in 2030. So, we've got more
manufacturing coming out. And we've gone
deep into why this is happening. A lot
of this has to do with the CHIPS Act,
which Donald Trump secretly hates. Well,
actually publicly hates, but secretly
sort of coordinates money on. Uh they
get investment tax credits, not only
from the chips act, 35% credits, but
also from the state of New York. They've
got multiple different segments over
here. But some things that are worth
pointing out when I look at this stock
tab over on the uh in the mechan
membership is uh that the two components
that are outside of data centers are
actually falling right now. Mobile and
client business unit sector and the
automotive and embedded business units
are actually seeing declines in memory
chip demand. Whereas really it's just
that AI cycle that's driving these price
demands. And I made this analogy on
average selling prices this morning
because these average selling prices are
going crazy. And I'll tell you the
analogy in just a moment that I talked
about uh to course members. But look at
this. This is the highlight version of
the uh average selling prices. We have
seen about 8 to 9% quarterly growth for
actual unit deliveries. Uh that works
out to about 32 to call it 40% unit
growth. So they are selling more units,
right? But they raised prices between 60
to 70% just from the first quarter to
the second quarter. And year-over-year
they raised prices twofold. They doubled
prices right here. Year-over-year
compared to the second quarter of last
year. DRAM prices are up 110%.
Roughly closer maybe to 115%. and nan
prices increased slightly more than
double. These are massive, massive
increases in price, which is exactly the
opposite of what happened in 2022 when
we saw memory chip prices crater,
absolutely collapse. It got so bad that
the dude who blogs about memory prices
basically said, "Man, it's so bad we
might actually end up having another
downturn in 2025, which was exactly the
opposite of what happened thanks to
artificial intelligence." And I'm not
trying to blame them or bag on them. You
know, AI changed a lot. But the point
is, people are so used to the boom and
bust cycles that they draw lines like
this and they're like, "It's going to be
bad again, boys and girls. It's going to
suck again." except what actually
happened. A boom in prices again. And
that's why I want to show you the
difference between these and then I'm
going to give you the analogy from this
morning's course member live stream.
Look at this. This uh was the last
quarterly report without the
highlighting. See right here, you can
see that mid 110% range and slightly
more than 100% increase in average
selling prices. Are you ready to see
their most recent earnings report, which
was an absolute banger? Like literal
banger. They beat on everything.
Absolutely smoking great earnings. The
best freaking earnings you could see. I
mean, this company is printing freaking
money. Okay. Why? Why in part are they
printing freaking money? Okay, you ready
for this? You think it's because the
business somehow is able to make a
better chip?
Maybe. I actually think it has a whole
lot more to do with this. Ready? Watch
when I flip this page. Boom.
Year-over-year,
prices have now increased. Instead of
110% 115% range, they've actually
increased quote 260%
increase in average selling prices. And
NAND products are up about 315%.
What? Yeah, dude. That's a triple and a
quadruple in average selling prices. So,
literally in the last report, I'm like,
"Oh my gosh, most of their increase in
revenue is coming because of price
increases because they're doubling the
prices of chips." Well, now they've
said, "Hey, JK, we're actually tripling
and quadrupling prices and that's why
we're making so much money." And this is
where it's really useful to understand
an analogy that I'm about to give you
because it helps you understand why
Micron is falling since my micron
warning and SanDisk is falling since my
micron warning. Why? Okay, so I'm going
to take out the good old whiteboard
here. And this I I actually I really
like this analogy. I hope you like it,
too. The uh the thing is when when I
make these analogies, sometimes people
sort of think, "All right, video's
over." They bail, and then they miss the
most important parts. These analogies, I
feel like, are the things that could
help you get that sort of aha moment.
You go, "Ah, I get it. Yeah, that makes
a lot of sense." And what I'm about to
tell you is really a way to think about
the entire artificial intelligence
euphoria that we've been seeing. Okay?
So the analogy I'm going to give you is
an analogy on real estate, but again,
we're using real estate just to help you
think of the chip sector. And you might
be like, Kevin, what do houses have to
do with chips? Watch this. Okay, so you
might know this. I run this company
called House Hack, aka Reinvest. See, I
got the little shirt on. Uh, and it's a
company that owns over 80, 90ish million
dollars in real estate and assets, no
bank debt. And, uh, you know, I think
we're doing really good things. We've
got cool apps. We got cool software, a
great software business on top of we own
real estate. And if somebody came to me
and said, "Hey, Kevin, we want you to
take a billion dollars and go buy
houses, but we want you to buy houses
the way you always buy houses. We want
you to get houses at a 20% discount. We
want you to play like Warren Buffett,
okay? Do your best to go out there, take
that billion bucks and get 20% discounts
on as many houses as you can. Work your
freaking butt off. It's going to be a
lot of work. Go make that ROI, right? In
this case, I need you to know that ROI
or a return on investment is going to
equal time. It would probably take me a
couple years to go deploy a billion
dollars. I think I would do a really
good job at it. Right? Okay. I'm telling
you, there's an analogy on chips in
this. I know people are like, "Kevin,
are you just trying to sell house hack?"
No, we're reinvest. Oh, no. No. It's an
analogy here. Ready? And this is why you
you stay to the ends of the videos. I
think you actually benefit here, but
with the aha moments. Are you ready for
this? Now, what if somebody came to me
and then said, "Kevin, we need you to
buy a billion dollars of real estate. We
don't actually care about the 20% or
ROI. We need this spent now and uh we're
not pointing a gun at your head. Uh if
you don't spend that billion dollars
now, call it within the next week or
two, we're going to kill you.
Okay,
guess what I'm going to do when it comes
to buying a billion dollars worth of
houses? Well, I'm going to go out there
and I'm going to buy every freaking
active listing and I don't care how much
I have to pay. I will overpay. I will
over bid. I will do whatever the hell I
need to do to get those houses cuz I
don't want to, you know. Okay. Why does
that matter relative to chips? Because
what's happening in the chip industry
right now is nobody cares about ROI.
People just care about I need it now. I
need the memory chips now. I need the
fuel cells now. We don't have time to
wait for the company FCL who just raised
$49 million of financing and for some
reason is memeing a little bit which is
nonsense because they actually need to
prove that they could sell their
product. Uh but you know more relatively
you could look at a company like Bloom
Energy. Bloom Energy kills it with fuel
cells. fuel cells. You could get a
permit for these in like 2 or 3 months
and pop them up. They're not anywhere
near as efficient as a turbine natural
gas power plant, but permits for that
take 4 to 5 years. They're certainly not
as cost effective as doing a turbine
natural gas power plant combined with
solar and wind. But that takes years.
Nobody wants to wait years for their
energy at a data center right now. They
want it now. They want their memory
chips now. They want their Nvidia GPUs.
Now, nobody cares about the ROI
and savvy investors realize that we
might be getting to the point of
insanity in terms of how much companies
are able to raise prices when they can
literally write in their reports, hey,
we just uh triple and quadrupled prices.
When a company tells you, hey, we're
dominating because we just tripled and
quadrupled prices. and you know it is a
cyclical commodity. People start going,
"Okay, how much better can this really
get?"
Maybe, just maybe, it's time to take a
little bit of profit.
Anyway, thanks for watching. Consider
subscribing to the channel. Make sure
you use that coupon code pope for more
perspectives uh and analysis,
fundamental analysis, technical
analysis, short-term trades, medium-term
trades, long-term trades, uh top stocks
that I'm looking at buying over the next
10 years and more. So, go check it out
over at meetke.com. And remember, if you
want that Reinvest AI, you can check
that out as well at houseack.com or
reinvest.co. Same thing. Thanks so much
for watching. We'll see you in the next
one. Goodbye.
>> Why not advertise these things that you
told us here? I feel like nobody else
knows about this.
>> We'll we'll try a little advertising and
see how it goes.
>> Congratulations, man. You have done so
much. People love you. People look up to
you.
>> Kevin Praath there, financial analyst
and YouTuber. Meet Kevin. Always great
to get your take.