저자: Dylan·Vinny · 외국인
요약 보기
테슬라 주가 급락 후 AI 버블 붕괴 우려를 제기. 테슬라 매출총이익률 하락·잉여현금흐름 악화, 로보택시 사업 수익성 의문과 자본지출 과다를 근거로 비관. 마이클 버리의 엔비디아·마이크론·PLTR·SOXX 숏 포지션을 소개하며 AI 데이터센터 수요 둔화 가능성 언급. 중국 오픈모델 확산으로 미국 AI 토큰 트래픽 이동 우려. 우주사업(SpaceX) 락업 해제(8/6) 부담 언급. 전반적으로 AI 관련 고평가 조정 가능성에 부정적 시각.
다듬은 글
Hello, welcome back to "Two Super Guys for Stocks." I'm Dylan. And I'm Vinny. So, please be patient. Originally, this was going to be a video about Tesla and when I was going to buy one. We try to find companies whose shares have fallen, like SLB, which has risen by 12%. We look for companies that have been unfairly affected but have good prospects for long-term growth.
This is the foundation of our channel. So I looked at Tesla, because a 20% drop in two days is a lot. That's a lot. I never rated Tesla very highly, so I thought: why not figure it out . This has escalated into the idea that the AI bubble has just burst. It seems to be happening right now. Just please, listen to me.
I understand how that sounds. That's exactly what it sounds like. I understand. Listen to my arguments. Things are happening that I do n't think people are talking about, and it's just getting started. For completeness: I wanted to find a price to buy Tesla shares. I do n't want to buy them. If you ask when to buy a Tesla—I don't know, maybe $80, and then I'll take a look.
Everything is too bad. And I'll explain to you exactly why. Yes, you and I have historically been wrong about Tesla, because even when we think it's overvalued, we're still wrong because the price remains high. Here is the QQQ index. It seems like we are going into decline. Something is happening here.
Okay, but of course, of course. You know, stocks can't go up forever, but I want to look at this. Pause. Yes, just a little consolidation. Just look at these assets. This is QQQ and their weight: Nvidia 7.58%, MU after this rally 5.63%, Tesla 3.29%. And, more importantly, SpaceX—1.5%, or sorry, 1.2%.
My mistake. Before we continue, I want to show you some of Michael Burry's deals. Yes, he constantly opens short positions. I understand. Yes, he was wrong in 2022. But now he's really making money from shorts. OK. Okay, about PLTR —he was absolutely right. He opened a short position in Tesla and said he was not going to close it.
This is stated in this publication. He says, "No, it still has a long way to go." Yeah, that's a bold move, dude. No, I'm not closing anything. Okay, he still has a short position on PLTR . He has a short position in MU. He increased his stake in Nvidia again. Short again short on Nvidia. and he also opened a short position on SOXX , which is a semiconductor ETF in the Nasdaq 100.
So he was pretty right. He was quite right . OK. Um, and he's not he's not closing these deals. I mean, he added. So the thing is, I understand. He is like a broken clock. You know, I, I, I understand. But it seems pretty accurate on many of them, except Nvidia, which is the most important. Yes. Right.
Um, we'll take a quick look at Tesla because something strange is happening there. When you have a company with a P/E ratio of a billion, well, not a billion, it's just very high. It's also 3.3% of the Nasdaq 100 index, and that's a problem. OK. OK. Um, now here, let me move my face so you can see the volume .
That's quite a big red stripe sticking out behind you, my friend. I thought you were just a little excited for a second. Yes. No, this is definitely a day of fall, man. This is not good. Again, I made the first half of this video when I wasn't looking at their fundamentals yet. Um, so those were the starting points for buying where I thought, you know, just based on the chart, I would make trades here, but frankly, I don't want to touch that.
Um, I just think she has a long way to go, but those would be interesting areas. Now here's why they fell so much. Here is one. Sorry. That's one of the reasons they fell so much . So you can see that their gross margin has gone down, I think, to 16.1 or 16.3. She was significantly taller all this time.
Let me turn my arrow. Much higher than that. And here she is just on a steep decline. OK. Hmm, a steep decline. Sorry, downward trend. It's too late. Hmm, that's not good. Well, it's not terrible, but it's really about their costs. Here is their free cash flow and operating cash flow. You usually don't want to see this.
Hmm, that's a bit of a problem. Let's continue. Wait. Here is their P/E ratio. Now I , I understand. You can't apply P/E to Elon Musk. I understand. I completely understand that. At the same time, when the indicator is in the hundreds and the company is growing rapidly—that's one thing, but you no longer are a rapidly growing company.
They are making very, very expensive bets as if that is still the case. But these bets won't pay off for many years to come, which I will also prove. Good. By the way, this figure of 289 is now 341 after, after this was not updated. So, after the last 20% drop, their profits are 341. Remember, this is 3% of the QQQ index.
So if these stocks fall another 50%. This will affect, you see how it happens with QQQ, right? Yes. This is an overestimation. I won't lie. This is wild. They make three huge bets. Good. Optimus robots. I'm not...okay. Robotaxi , which I'll talk about, and physical AI, which I'll talk about now. Robotaxis.
Do you remember how long he's been talking about this? I don't know, maybe since 2015. Yes. It seems like we've looked at this before . I think it was in 2017 or thereabouts . Um, that's the thing. When you talk to Tesla fans or analysts on YouTube, they think robo-taxis will generate billions upon billions of dollars in revenue.
I don't know where they get these numbers from. This is similar to Katie Wood's theories, but Waymo, which has been doing robotaxi for a long time, guys...Well, I, I'm in Phoenix. I've had this since 2020. I've been using this for six years now. It is operating at almost full capacity. This is not news.
OK. They, they're still not profitable. The Waymo segment itself is still not profitable. And now if you look at the Google spreadsheet for other rates—it's different , but just Waymo... Waymo is getting close to profitability on its own, but remember, look at the fleet. So, total vehicles in operation: 3,871.
Tesla has 42. And Tesla says it's going to do this starting in 2017. Good. They don't talk about the number of trips because they have 42 cars and they probably don't make any trips. But Waymo does 500,000 trips a week and has 221 million driverless miles. I think they included this in their latest report because Tesla would count miles, not trips.
That's my theory. Um, okay. The Tesla only has 380 unattended miles, and we don't know on what trips. So it looks like they're trying to get into a certain area, but they're not even close to there. It won't make them billions upon billions of dollars like everyone thinks, but Tesla itself is spending money as if it will.
So, here are their capital expenditures for the entire year 2026. I know, I'm still all about Tesla. Stay with me. Um, here's their capital expenditures for 2026. Over 25 billion. They have a cash reserve of 44.1 billion. That's pretty good. That's a lot of money. I understand. They are increasing their capital expenditures by 142%, just like all other companies.
But there is a big difference between Google, Meta, and all of these companies. That's the big difference. Okay, this is their quarterly net profit. Now this is a company that is going to spend 25, how much is there? 25 billion for capital expenditures. And these are their profit figures. This doesn't add up at all.
The pun is intentional. Boom. Compared to Google, oh, that's $ 45 billion in capital expenditure, $120 billion in revenue? Yeah, that's not even close. 34% operating margin, it seems. So like, yeah, we're talking about $40 billion sometimes in operating profit. They can afford $45 billion in capital.
Yes. Yes. This is not the same thing . They absolutely cannot . I mean, in Q1 and Q2 they made 477 million , where the money will run out. Remember , their margin ratio on cars is falling. They are not going to earn more. They are making less and less money. Well, it's not that it's less and less, but they're not making enough money.
Subscription revenue. But yes, the average selling price of their cars is going down. They used to sell cars for $70,000 . Now they sell cars for $50,000 . And the state benefits tied to them are also no longer there. Here too, look, look, look at Q3, Q4, Q1: 1.3, 844, 77. What's going on? Well, um, as it's called, green energy credits are disappearing for them too.
So, regulatory credits—that's what they call them. Aha. High returns , but they are not good, it is not good , that is what it means; now you see their operating margins have collapsed to 1.4%–4%. This is largely due to operating expenses, as well as what you just said about loans. Now there are some more weird things, but remember that their three big bets are Optimus robots.
Sorry. Give me 15 years and then I'll think about it. I don't... I think it's like the robo-taxis he talked about in 2017, although Waymo actually launched in 2020. This is different. Um, that's not even remotely similar. I do n't even know what the market is for this right now . And secondly, robo-taxis will not be profitable in any way.
And you will have to spend a lot of money on it. This won't work for much longer. Maybe 15 years old again. They spend money as if they are going to earn simply crazy amounts of money. Yes. I think their path to robotaxis is shorter, because the cost of one car will be lower than Waymo's, because Waymo buys third-party cars and invests 100 thousand in sensors.
Um, because of the camera-only system that Tesla uses. It will be a cheaper car with higher margins, but it will still take time to scale. That's how it should be. But they also said, I think in 2021, that it was like, "Oh, yeah. Three months—and that's it." Of course. We have 42 cars. So, yes. OK. Um, there's something else...
Oh, [ __ ]. I pressed the wrong button. Wait. Sorry. Sorry. Sorry. Uh, okay. SpaceX. So, this is 1.2% of Qs. Again, not related to Tesla, but just adding percentage points. OK. So, we have 4.7% of Tesla and SpaceX. SpaceX shares are falling rapidly. They fell from 215 to about 113. In addition, the lockdown period, the first part of it— it has a very unique period that is broken down into stages.
Ends August 6th. I don't see any way this could increase after August 6th. Agree. This is not good. OK. Yeah, that doesn't... It sounds like a problem. Now it all moves with chips. If you look at Micron and Nvidia, that's 13% of Qs in just those two companies. But what people don't talk about is that Meta and XAI have excess computing power, which shouldn't be the case in a world where computing is considered the main constraint.
Okay, that means one of two things. To be fair , the argument against this is that when you train a bunch of models, it happens in batches , and then you're left with excess computing power, and that may not always be the case. So then they can go back to studying. So I understand this argument. I understand.
However, at the same time, AI models account for approximately 30 to 58% of token-based traffic in the US. This is the Alibaba model. This is Deepseek and a few others. Okay. That was, what was that? 4%. Yes. at the beginning of 2025. So, that's approaching 60% of the traffic from US customers that is token-based.
They are switching to Chinese models because it is much cheaper. At the same time, there's Anthropic and all these other companies that are spending crazy money and relying on American companies to use their products. And China just says, "We'll take care of it, man. Don't worry." This is a problem.
This is just madness. I mean, Google talked about resource constraints . Alibaba is limited in resources. Meta is resource-constrained, and they did mention the possibility of launching their own cloud service. But yeah, it's unusual for this field not to have cloud computing right now. Yes, I think as China continues to release more and more models for open access, combined with giants like Tesla, whose valuations are incredibly high, it could lead to a market crash.
If Nvidia starts to fall and Michael Bury turns out to be right, it won't be fun at all . This will be difficult. I really think Michael Bury might be right about Nvidia. I just don't see any prospects. He has a great article on how it all works— cash circulation. None of this is realized profit. It's all about financing, it's interesting.
I really think this is going to be a problem. I don't know about you. I know, it sounds a bit like a conspiracy theory about the end of the world. I understand. No, that's interesting. I'm trying to remember the name of the company responsible for the conference call that is believed to have caused the dot-com bubble to burst.
Is this WorldCom? There is one particular company that people are blaming for their conference call being the beginning of the end. I can't remember off the top of my head who it was. I think so , but there is such a narrative. So you never know. One of these reporting conferences could be what ultimately triggers a major, so to speak, price...I don't even know what to call it.
... a pullback or a crash, a bubble burst, let's say a bubble burst. Yes , of course. It is possible. You don't know who it will be, but......yes, I know, or when, right ? It can also be an incredibly slow , tedious process, which it usually is. I just...I don't see us going back to the highs. Let me take a quick look at the graph.
I would be very, very surprised if we got close to 750 again. Also, oil is getting more expensive again. This whole conflict in Iran. It's not going anywhere, we're already near the $100 mark . We are approaching $100 per barrel. If not, then we're already there again. These, I mean, are all problems.
By the way, Michael Berry opened a short position on Caterpillar. I forgot about that. It's like a big deal around data centers for AI. Yes, Caterpillar. Well, Syria and Iraq have said they are going to work together to build an oil pipeline so that we can bypass the Strait of Hormuz by 2032. I don't know what year they listed there, but I'm sure it's not soon.
We'll have to wait a bit . Um, yeah, let us know what you think. But in my opinion, Google's profits are falling due to capital expenditures. At Tesla, the reports show that they're not actually making money, they're spending money, and I do n't know where it's coming from, except from their cash reserves, I guess.
Hmm, I don't see this ending well. And if MU and Nvidia start to fall and companies continue to use Chinese models for business, that's just... that's not good. To me, this makes no sense at all . But maybe someone can explain it. Although maybe it's good for our Baba shares. It should be. I hope I bought more. Good. Bye, guys. Until next time.
Comments 0