---
title: 'the *uckening'
source: 'https://youtube.com/watch?v=78Xg31Tina0'
video_id: '78Xg31Tina0'
date: 2026-07-24
duration_sec: 657
channel: 'Meet Kevin'
---

# the *uckening

> Source: [the *uckening](https://youtube.com/watch?v=78Xg31Tina0)

## Summary

This video analyzes the current market downturn, attributing it to massive capital expenditures by big tech companies on AI infrastructure, particularly the rising cost of high-bandwidth memory (HBM). The host introduces a subscriber's thesis that the inflection point of memory prices could signal a peak in capex spending, potentially allowing tech stocks to recover.

### Key Points

- **Market Context** [00:02] — VIX low under 19, Nasdaq 100 showing lower lows and lower highs, broken uptrend wedge. S&P 500 near all-time highs, indicating a bifurcated market.
- **Google's Capital Raise** [01:55] — Google raised $80 billion total, with $40 billion in convertible notes. The raise roughly marked the stock top. Another $40 billion at-the-money facility remains unsold.
- **AI Spenders Under Pressure** [04:40] — Companies spending heavily on AI (Google, Microsoft, Meta) are underperforming. Apple, which isn't spending as much, is up 18% year-to-date. Microsoft and Meta are down 21% and 8% respectively.
- **Subscriber Thesis: Memory Prices** [06:27] — Subscriber suggests that larger context windows in AI require more HBM, causing prices to skyrocket 370% YoY. This is inflating capex. AI is not getting smarter, just remembering better.
- **Microsoft's Capex Breakdown** [08:17] — Microsoft stated 25% of their capex spend is solely due to memory price increases. This highlights the impact of HBM costs on overall spending.
- **Indicator for Capex Peak** [08:47] — When the growth rate of memory prices stops inflecting up, hyperscalers' capex may top out. This could be a leading indicator for stock recovery.

### Conclusion

The video suggests that monitoring high-bandwidth memory price trends could provide an early signal for when the AI capex cycle peaks, potentially allowing tech stocks to stabilize and recover.

## Transcript

happen in June. We talked about this regarding SpaceX and Google, and obviously markets have basically topped since then. The question is, when does it end? And the best way to understand when it ends, in my opinion, is to
understand what's causing the pain. And I want to shout out somebody, I don't know if they want me to use their name, but somebody emailed me this idea. Uh it. It has to do with memory. We're going to talk about that in just a
moment, but let's set up the baseline first. So, the first thing that we have to know is even though the market volatility index for certainly the Dow, the VIX here is very low. We're under 19. Obviously, we have seen a topping on
the Nasdaq 100, which we like to use the Qs to look at uh on a regular basis. We're seeing lower lows, although we've kind of shelved around 690. We're consistently seeing lower lows, and we're seeing lower highs over here. Not
fantastic. We've broken this uh mini little uptrend wedge over here, which is a pretty common technical pattern that when you've got this sort of converging downside. The question is, where is the floor? Does it actually stay at 690, or
do we go down to 675? We'll find out. Part of what we talk about here will help us address this. Now, a larger issue is not that the entire market is selling off. If we look at a stock like RSP, which is the S&P
500 equal weight, we can see we're within a couple like two or three cares? You don't even have to go to the equal weight. Frankly, you can go to the regular spy or spy M over here, and we're bobbing near all-time highs. This
is really what the market is experiencing, a frustration with capital expenditures. Uh and the way I like to analyze this is by looking at Google and what Google actually said in their earnings report.
Now, I broke this down in like nearly a 50-minute video yesterday on Google, so again. You can watch that video, but I'll give you a quick bottom-line note. Google raised a convertible round, an equity round, the Berkshire underpinned
round, right? Where Greg Abel got like a 7 or 8% discount and he's in it like 350 on Google. It's trading for like 320 right now. Uh in total, they raised about $80 billion. In their earnings note yesterday, they said $40 billion of
convertible because that hasn't converted yet, but $40 billion has not hit the market yet. So, maybe somewhere around 22-ish, 18 to 22-ish billion of the raise hit the market, and it really roughly marked the top of the stock.
roughly marked the top of the stock. They announced the raise on uh June 1st is when they announced the raise, which is right here. Google stock was uh closed at 372 that day. They hit 381 before that raise announcement. So, they
before that raise announcement. So, they were at 381. Google was actually on like experiencing this uptrend velocity because they were cash flow positive and their balance sheet was starting to run low on available cash. They had a lot of
bills to pay, but they had enough cash to cover it assuming their expenses didn't keep skyrocketing and they didn't go cash flow negative, right? Obviously negative, and so they needed to raise money, hence why they raised money.
Somebody did leave a comment yesterday and and I was kind of like, "Ah, something out of context again." And I always like to give you the context and Somebody said, "Oh, you know, well, Kevin said Google's going to 500." And I
just want to clarify what I said because I went back and looked. What I said is Google's momentum and a break through at 400 could lead it to go to 500 on only a technical basis in the near term unless obviously
something comes in to stop that. And how about an $80 billion money raise? trying to say, oh, you know, every look at technical calls, usually when we've got this momentum like we've had
at Google, the earnings that they had were really just propelling them on April 29th. They were actually doing fantastic. The money raise, the suckening in my opinion, marked a moment. And now, I
clear the record. I bring it up because it's something you as an investor should pay attention to. Look at what's happening. Companies are going up that are benefiting from AI as long as they don't have to go to the market and raise
debt like Meta or raise a debt and equity like Google or spend more than they're making, which Microsoft is getting pretty close to that. So, if we that as well. These are the companies that are getting hammered. The ones that
are spending on that AI trade are the ones getting beat up. Now, that's obvious at this point, right? We know that. The big point I want to make here, one of them, and then I'm going to get into what what a subscriber sent to me.
The big point I want to start with is Google hasn't even started with their 40 billion dollar sale. Yeah, in their last note, they said as of June 30th, we have not sold any of our 40 billion dollar at the money facility. So, more money raise
is still coming. Okay. Now, part of the reason that money that we want to pay attention to and maybe gives us a tool for finding a start like the hyper scalers will start actually rotating up again. Because
right now they're rotating down, okay? Look at the year-to-date performance of Google, Microsoft, and Meta versus Apple. Right? Apple is there's not a lot of upside in my opinion in Apple stock price at its
current valuation, but it's got momentum. Why does it have momentum to the upside right now? At 320 bucks, you know, it's been stuck at like 275 forever. Uh, because they're not spending like drunken sailors like these
others. Apple's up 18% year-to-date. Google's up 1% year-to-date. Microsoft and Meta are down 21 and 8%. Now, look at just the last 1 month and you've got Apple, Microsoft, and Meta up a little bit off the floors, but you could see
the downtrend on this last month. It all peaked right around here. Which what happened right literally around here? SpaceX topped. >> Which was your second great suckening. So, it's like you had Google and then
you had the SpaceX suckening. Point is, the more of these issuances we get, so like this Anthropic IPO that's likely coming, is probably going to lead to even more down pressure, suckening money out of the market as Google also
sucks, right? Okay, fine. Now, what did the subscriber say? And this is kind of what I really wanted to to bring up. So, I think it's a very interesting thesis. Now, I want to be clear, it's a thesis. So, let me make sure I got through the
little note I wanted to make because I don't think I finished the thought yesterday. TSMC in their earnings call said that anything outside of AI is doing basically flat to negative. Worth noting, that's Taiwan Semi. The
only thing booming right now are AI chips. Everything else, toilet. Okay, so. The concept they emailed me, which I actually I you know, I'm I'm going to call them a pro for just bringing for
you can always send me concepts. If you want, you could even use, you know, news information is. We have a community tab. Post your ideas in there. Like, be a great place to send me ideas and
Somebody sent me an analysis and they said, "Kevin, read this. You'll find it interesting." And basically, they linked seen, like at Micron, where we've seen
memory prices skyrocket 370% year-over-year. Last quarter, that was like a 2x. Now, it's a 4x. So, memory prices skyrocketed. Why? The Anthropic moment, bigger context windows, great. My thesis is bigger context windows make
AI better, faster, but there's a limit to how much better they can get. Like AI didn't actually get smarter, they just remember better. Mm, really interesting. idea, okay? Think about that for a moment. Why is Why are these LLMs
not because they're fundamentally getting more intelligent. Think about future of AI in terms of like capping out. But, the concept that they brought up with is they linked together that Microsoft complained that about 25% of
their capex spend was solely due to memory prices going up. Which is interesting because what we're trying to find is when is the peak of capex spending. Google just raised their capex spending another 15 billion
dollars. A lot of it entirely possibly because of memory prices. So, what's the concern or what's the thing to look for? Maybe the thing to look for in terms these guys bottom out is after we get an Anthropic IPO and when the growth rate
of memory prices going up stops. I'm not saying memory prices coming down. I'm saying we stop seeing, oh, now we're up 200%. Oh, JK, now we're up 370% what high bandwidth memory is for these context windows. They're literally
generally stick in your motherboard like this and they're all separate with the bus plugged in the back, right? Assuming you've got like eight in there, usually people just have or two, right? Some have one. Um, now take all those chips
out, turn them on their sides, stack them and sandwich them together, drill a hole through the top of them, and now let memory run vertically through those DRAM chips as well as then into, you know, your actual uh
uh you know, chipset server rack, whatever, right? That's essentially high same crap sandwiched together and rearranged in a different way to where you could do more at the same time. That's holding your memory context. Good
for training, good for inference. Okay. Now, when those memory prices stop Now, when those memory prices stop inflecting up, then guess what happens? All the hyperscalers don't have to increase their capex spend solely
their capex spend might actually top out. So, what a top could be in the hardware cycle might be when memory prices stop inflecting up because then companies like Microsoft and Google don't have to raise their capex spend
up, and then those stocks can actually start rising. Remember, last point, Mark Zuck, he told us, "Hey, we're going to opportunistically sell excess compute." Stock freaking rallied on that because people are like, "Wow, you guys are
AI and AI now?" Huh. People are allergic to the spend and the debt right now, and they should be cuz it's risky. But, this This is the video to bookmark because I think it could be an early indicator for that top. Thanks
>> Why not advertise these things that you told us here? I feel like nobody else see how it goes. >> Congratulations, man. You have done look up to you. >> Kevin Paffrath, a financial analyst and
YouTuber, Meet Kevin. Always great to get your take.