저자: TomLee · 외국인
요약 보기
SpaceX IPO 전 기술주 매물 출회는 기관의 나스닥100 편입·비중 조정 수요에 따른 일시적 현상이며 기술 트레이드를 훼손하지 않는다고 봄. 반도체 메모리 주가가 전주 저점 위를 유지해 차트가 무너지지 않았다고 평가. 하이퍼스케일러의 설비투자 확대(잉여현금흐름 음전환)는 장기적으로 승자가 될 것이라 낙관하며, 마이크론·삼성·TSMC 언급을 인프라 장기 투자의 근거로 제시. 엔비디아의 데이터센터 프로젝트 금융 지원은 수요 약화가 아니라 AI 인프라 확장을 돕는 것이라 해석. 테슬라 Optimus 등 로봇 사업은 회의적으로 봄.
다듬은 글
Well, Scott, I think the market is certainly showing a lot of signs of nervousness. Uh , you know, the nervousness before the SpaceX IPO, and also just consolidating all those profits. Additionally, we know that Google, Meta, OpenAI, and Anthropic will attract a lot of capital in the future. So I think the market is trying to price this in ahead of the SpaceX IPO, which is happening this Friday.
However, I consider this a healthy phenomenon and I don't think it will actually disrupt the technology trade. What's happening, you know, with chipmaker stocks? Do you think that the selling there is most related to positioning ahead of the SpaceX IPO, which will take place at the end of the week? How would you assess what happened there?
Uh , yeah, you're right, Scott. I mean, SpaceX is going public at a $75 billion valuation, and because of the Nasdaq 100, many institutional funds will now have to raise money not only to participate in the IPO, but also to have cash to build positions, because they don't know if they'll only get 25% of the distribution if demand exceeds supply four times.
But then, to reach the full market weight of the position, they will have to buy a lot on the secondary market. I think this means selling some of the recent market leaders. A good sign , in my opinion, is that shares of memory companies have held above Friday's lows, the same goes for semiconductors.
So I think that even though there are noticeable bumps there, I don't think that means that the technical charts are completely broken. Does the market need SpaceX to trade well on its debut? Uh, yeah, I think that's important, Scott. We expect them to trade well because it is an important company and Elon Musk is a unique figure, but there is also a good plan for further growth and a very large target market.
If they trade poorly, I think that will lend weight to the arguments of many skeptics who, uh, they announced $25 billion in capital spending. I think they will have to spend much more. Currently, they have 30 billion in cash in their accounts . So it will take a lot of money and a lot of effort to do what they set out to do.
That's how complicated it is. Well, so it's really hard for us, I think the only thing that's going to happen, and this is my prediction, is that SpaceX will have to buy Tesla. They have so much synergy together, and it's just much easier for Elon to run both companies. And I think that's actually one of the reasons why people hold these stocks is because there's going to be a merger between the two companies, no matter how expensive they are.
Yeah, and I mean, he quite deliberately didn't give a lot of details on that, which I think is interesting, because that's the question we're all asking: How long or when will we see these companies come together, well, under one leader? But I mean, robotics is an interesting topic, and I will say that I'm a little surprised by the extent of the pressure today , because, as you said, this narrative was already clear.
I mean, it was a bullish thesis, or something that the bulls have at least been holding on to now, I mean, for many years now. So let's assume that robotics is, of course, a priority , then what proof do investors need to believe that Optimus can become a real business and not just another long-term promise.
They talked a lot about how difficult it was to create Optimus' arms and ensure they were dexterous. This is a big, big problem. And for me , I don't even think that the humanoid form factor is really the best way to build robots if you're trying to do different tasks. And will we all have robots running around the house, climbing stairs, etc.?
So again , you have to really believe in this long-term theme, that it will work and we will have millions of robots. I just don't see it happening. Okay, let's take a step back, you know, look at the big picture, more macroscopically, holistically. It seems like the key change happened about a quarter ago, when we got earnings reports from some of these mega-corporations: Alphabet, Microsoft, Meta on the same day.
They all announced increased capital spending. Alphabet, of course, just did this yesterday. But the big difference seems to be that now free cash flow has become negative. This is no longer a good thing, at least in terms of stock market reaction. AI infrastructure companies still seem to be responding positively.
Do you see a world where these two strategies— hyperscalers and infrastructure companies—can merge into one correlated trend again, or will this inversion remain, at least for a while? You rightly remarked. I live in a world where investing in innovation ultimately pays off. So when you see Google and Microsoft spending a ton of money on infrastructure, I think in the long run they're going to be big winners.
You just have to go through this cycle, right? The best part of yesterday's Tesla conference was the mention of Micron, Samsung, and TSMC, right? This truly confirms the fact that infrastructure investments are for the long term. They announced $25 billion in capital expenditures. This is very good for the entire chip and infrastructure sector.
And indeed, you shouldn't swim against the current. That's where the money goes . These are the companies that are doing well. Ultimately, if you believe in AI like we do, then this capital investment will be a fantastic thing for our economy and this investment. So I'm not worried about the long term .
On the contrary, I am even more optimistic about this. And I think hyperscalers that start trading at very attractive valuations will be big winners because they're investing in the right things. Yes, I completely agree. I really think that was the most brilliant moment of the report, not even necessarily related to Tesla, but how they promoted other technologies and their uses.
Every few years, the stock market creates a narrative that sounds compelling. This market narrative is simple. Chinese models are becoming cheaper, and many companies are reaching peak profitability. Shares of semiconductor companies have already doubled, leading many investors to believe that the AI boom is finally slowing down.
But here's the weird part. While investors wait, companies actually building AI are sending a very different signal. Microsoft, Amazon, Alphabet, Meta, Samsung and SK Hynix. One after another they say essentially the same thing . We still don't have enough infrastructure for AI. So today, let's put aside the headlines and look at reality to find out what's really going on.
Let's start with Microsoft's earnings, because they answered the one question investors have been asking all year. Can AI really generate enough money to justify hundreds of billions of dollars in spending? Microsoft's answer was simple: yes. Azure's growth accelerated, cloud adoption jumped again, and management explained : as soon as new AI capabilities become available, customers start using them almost immediately.
This means that Microsoft is not building empty data centers waiting for customers. Customers are already waiting for these capacities. That's a huge difference. Next comes Amazon. And this report was perhaps even more important. AWS has shown one of the strongest growth rates in recent years. Operating margins surprised analysts , and instead of slowing investment, Amazon again increased its capital spending plan.
Wall Street usually doesn't like to hear about higher costs. But this time, investors reacted positively. Why? Because Andy explained something very important to Jesse. Amazon doesn't predict future demand. Most of the infrastructure they build is backed by real customer commitments. In other words, they are expanding because customers have already booked capacity.
Not because they hope someone will eventually show up. Now think about what that means. The world's largest cloud providers are spending more money than ever before. And yet each of them says the same thing. We need even more infrastructure. This doesn't look like an industry that's preparing for a slowdown.
It looks like an industry that is barely keeping up with demand. This is where NVIDIA comes into the conversation. In recent weeks, you've probably seen headlines about NVIDIA supporting funding for large-scale AI data center projects . Some investors instantly panicked, arguing that Nvidia was forced to help customers because demand was weakening.
But this interpretation ignores something much more important. Imagine building a city. If a construction company suddenly can't get financing, the city stops growing. This does not mean that people do not need homes. It simply means that construction is slowing down. It seems like Nvidia is trying to prevent just such a situation.
They ensure that AI infrastructure continues to expand, instead of being held back by financial issues. That's a whole different story.
Comments 0