저자: TomLee · 외국인
요약 보기
기업들이 비용절감과 기술 활용으로 이익이 25% 성장한 반면 시장은 10% 상승해 주식이 상대적으로 저렴해졌다고 주장. AI 수요는 실재하며 AI 에이전트 보안 등 기술적 해법이 등장하고 있음. 엔비디아는 저평가 상태이며 목표가 대폭 상향 여지가 있다고 봄. 데이터센터 반대와 부채 조달 리스크(Oracle CDS 확대)를 언급했으나 알파벳은 재무 여력이 충분하다고 평가. 케이던스 디자인과 아날로그 디바이시스는 장기 보유 중인 종목으로 소개됨. SK하이닉스가 엔비디아보다 밸류에이션 매력이 크다는 비교 분석이 소개됨.
다듬은 글
I think people forget that they tend to look at the stock market as something that has grown significantly in 5 years, and many people just buy stocks without thinking, but they don't take into account that companies are using technology. They adapted to all these "black swans" and cut costs, so this year profits grew by 25%, and the market only by 10.
So the market became cheaper. And I think the point is that we trust our money to the CEOs of companies that have proven that they are trustworthy, because they don't just sit there and tell the board of directors that everything is fine. They are in crisis management mode and are constantly cutting costs.
You know, you're never going to be at a peak where people want a low P/E and earnings are rising unless it's a deeply cyclical sector. And the US economy is the complete opposite of a deep cyclical sector, right? We are now more of a low- volatility economy, so multipliers should increase significantly.
First assumption: Muse just proved that the demand for AI is real, but so is the resistance to data centers. So let's invite Bell Point Asset Management's Chief Market Strategist, David Nelson. David, by the way, you wrote another great piece contrasting the opposition to data centers with the huge demand for Muse, and I want to get to that, but there was other news this morning that was related to that.
Jensen took a step forward. He has a development that he calls the Agent Security Platform. He says it will contain and control AI agents, which will be given minimal rights. Could this slow down some of the fear mongering we're talking about? I would really hope for that. Most hyperscalers have reported incidents in the past few weeks.
So the threat exists, and the fear that accompanies it is a reality. Nvidia, he said, had just announced open- source software that monitors the behavior of agents and can isolate them if they stray beyond the bounds of what is permitted. They claim it can be done in milliseconds, and the good news is that it's not just Nvidia's proprietary technology.
It will be available for Arm and Intel. For me, the key takeaway is that the solution to these threats is more technology, more AI and development , not less. I always thought so. If, to whatever extent, this is a real threat, I think commerce and capitalism will respond, because there is good money to be made from it.
By the way, Nvidia was one of your favorite stocks in August. Do you still hold these positions? Yes, I still have them . They announced a share buyback, but that's not the reason I own them. You evaluate it in terms of capitalization versus growth rate. They are priced incorrectly. It is installed completely incorrectly.
Today they should cost about 15-20% more. 150 billion will add another 235 billion between now and 2028. I mean, this is a real " money printing machine." Regarding your comment, the shares should be worth significantly more. OK. So, this weekend, Mark Zuckerberg announced that within five years, every person on the planet will have a personal AI that knows their life, understands their goals, and does their work for them .
That sounds great. Who could oppose this? How do you, as an investor, look at this? Well, I tried to think and ask myself what AI will be like in five years. And I think we're moving towards the agent becoming the interface and the application becoming the " infrastructure." So in the future, you won't be opening Expedia on your computer.
Your agent will contact her himself. Meta has now provided what I consider to be very good intelligence available to 3.5 billion people. So trust will likely be a crucial factor in the development of this technology. Who has it and who doesn't—that's what will separate the winners from the losers. So, speaking of trust, you mentioned the scale of the opposition, and that could be a topic in the midterm elections, but I found it interesting that you said that a bigger obstacle than politics could be bond yields and the need to raise debt.
Of course, we saw what happened with Oracle, and the Jupiter situation in New Mexico. Their CDS spreads skyrocketed. So do you avoid companies that need to raise money? Because it seems like they're all going public now. They come out, but to varying degrees. If you look at the difference between Oracle and...I don't presume to determine who will win, because it may be a successful strategy, but the risk here is much greater.
You're talking about the widening of CDS spreads, which have increased significantly over the last year, unlike some hyperscalers. I look at a company like Alphabet. Alphabet could pay off all its debt and still be left with a hundred billion. So they have significant funds, and that could be a deciding factor in the coming years for who wins this big race.
But the main takeaway for you is that so many people have picked up on it, they like it, they're already using it, and that shows that AI will be accepted, it will win and it will live up to expectations, even though we're not 100% sure who exactly will be the winner. I don't know exactly who will win, but I can say that AI and the market will win because the consumer will win.
So the political rhetoric that we're hearing now, I think, will die down shortly after the midterm elections, and those governors and senators who opposed it will probably gradually change their position because they don't want to lose the taxes that will come in. True. It's economic activity and the very fact that their voters want these products.
Let's quickly go through some stock picks. You've been very accurate lately. You like Cadence Design now. Cadence is showing confidence but has made a series of lower highs. It is currently at a critical point of support and resistance. Why do you like her today? Well, it probably feels a little better than most software companies today.
Software is currently taking a hit for other reasons. Almost every modern chip, and the semiconductor cycle is still on the rise. Almost every modern chip in the world is developed using software from one of two companies. This is one of them. Cadence is the larger of them. They sell software to engineers to design these very complex things.
That's right. complex chips. So, I hate this schedule. The schedule is just terrible. There is no doubt about it. But I looked, it fell 22% from its highs. This is a survivor company. This company is not experiencing as much of a shake-up as some others. Another name that has faced turmoil in the past and emerged victorious, Analog Devices , again, had great movement, but has actually been trending downward since the summer.
Although if we look at the September movement, it is quite impressive. It looks like the worst may be behind us. You know, Analog is a more diversified way to invest in the technology cycle and certainly in the chip manufacturing cycle. They are going to make money in the field of AI. From 13 to 20% of their business is in data centers, but this is the physical world.
They work with factories, cars, hospitals, power grids. Anything that uses a chip is likely to be related to ADI in some way. So, I own these shares. We have owned them for a very long time. I don't plan on getting rid of them anytime soon. OK. I'll also note for the audience that you also like MetLife.
Jensen Huang's personal fortune has crossed the $200 billion mark. The Nvidia chip smuggling case continues to grow. And there's a strong argument that the best AI stock right now isn't even Nvidia. Let's analyze three stories from today. Jensen Huang's personal fortune today surpassed $200 billion, reaching $203.2 billion, making him the seventh richest person on the planet.
The reason is simple. Nvidia shares set a new record today, rising nearly 2% , bringing the company's market cap closer to the $6 trillion mark we've been tracking all week. This growth is fueled by two specific factors. Optimism around Meta's new AI agent Muse and Nvidia's $150 billion share buyback program expansion .
And fundamental indicators confirm this. Nvidia now controls about 80% of the advanced AI chip market and generated $58.3 billion in quarterly revenue, up 211% year-over-year . When your company grows like this, your 3.5% founder's share grows with it. Second, a quick update on yesterday's story: the arrest of Greg Louis, a California business owner accused of smuggling over $300 million worth of export-controlled Nvidia servers into China via Malaysia and Singapore .
More publications reported on this today , and it's worth reiterating why it's important. This is at least the third major chip smuggling case this year involving Nvidia equipment, following charges against a Super Micro co-founder in March and a separate case in Taiwan. This pattern becomes impossible to ignore .
The same shortage that is driving Nvidia shares to record highs is forcing people to take legal risks to smuggle to circumvent export bans. Third, and this is a discussion worth pondering: a new comparison proves that SK Hynix may be a better investment in AI than Nvidia right now. Nvidia's numbers are certainly stronger on paper: 55.6% net margin and $96.7 billion in free cash flow, built on 65.5% revenue growth last year.
SK Hynix showed a still impressive 44.2% net margin with $18.2 billion in free cash flow. The real debate is not about who is more profitable , but about the assessment of value. SK Hynix, which supplies high-speed memory for every AI server, trades at a much lower multiple than Nvidia, despite its high profitability.
For investors who fear that Nvidia's price has already priced in years of future growth, SK Hynix is being offered as a way to bet on the same AI infrastructure boom with a much cheaper entry point. So, today, Huang's personal fortune has just exceeded $200 billion. As the fight against smuggling continues to expand, Wall Street is starting to ask whether a smarter investment in AI might be in the memory sector, further down the supply chain, rather than GPUs. Let me know whose side you are on in this debate.
Comments 0