저자: TomLee · 외국인
요약 보기
Tom Lee은 중간선거 이후 AI 관련주가 데이터센터 정책 우호로 반등하고, 고용 둔화로 인플레가 완화되어 10년물 금리가 5% 아래로 내려갈 것이라 본다. S&P500은 연말 8200~8400 수준을 예상하며, 10월 중순~중간선거 전 약세 가능성은 언급하되 강한 하락은 제한적이라 본다. 소프트웨어 및 AI-first 기업의 성장과 빅테크 설비투자 지속을 강조한다. 후반부에는 Nvidia 주문 증가와 ARK의 Nvidia 매수·AMD 매도 흐름을 전달하며 투자 조언이 아님을 밝힌다.
다듬은 글
AI and cryptocurrency stocks have been dragged down by the midterm elections and the perception that Democrats do not support data centers and the Clarity Act. But our position is that after the election, Democrats will start supporting data centers and the Transparency Act because, first, such centers usually provide significant economic benefits to communities, and the law had many supporters among Democrats.
So I think AI stocks will benefit after the election because changes in Congress will actually give the green light to some projects. You, Tom, are interested in the employment report, which you believe indicates that the previous August report was somewhat off the mark, and from this you conclude that inflationary pressures from the labor market are unlikely.
So you're back to the idea that if you can control oil prices, inflation might stop being a problem. Yes, good morning, Joe. That's right. I think that on the eve of last week's employment report, the market expected the Fed to be as hawkish as possible. One that included three rate hikes, including a 75% probability of an increase in October.
This employment report was weaker. In two weeks we will receive the inflation report for September. In our opinion, yes. We believe that one-off factors are fading, so we expect softer inflation figures over the next 6 months. This will not only allow the Fed to retreat from its hawkish policy, but also, in my opinion, will contribute to the normalization of bond yields as inflation uncertainty begins to subside.
I see a lot of clues in this, as tech stocks are hitting new highs and cryptocurrencies are performing well —they wouldn't be able to perform so well under tighter monetary policy. So I think they expect easing financial conditions ahead. So, you read the " signs" in different markets to understand what's going on.
That would be significant, Tom. So when you say yields are normalizing, what do you think they should be for the 10- year bond right now if there weren't these, what you think, false signals in different markets? I think we can base our thinking on the fact that before Jackson Hole, the 10- year bond yield was 4.75%.
And I believe that any indicator below 5% will be perceived by the markets as a very positive signal for risky assets. And I think there is a good chance we will see yields below 5% in the next 6 months. You also expect the S&P index to rise by the end of the year , perhaps by as much as 9%. So, you're predicting somewhere around 8200, maybe 8400.
And you're calling the current market something that, in my opinion , is again going against conventional wisdom. You think this market is like a compressed spring, ready to soar much higher. Almost like a containment mechanism , although most believe it is overheated and growing only thanks to AI. Uh, yes, I think we're living in unprecedented times because income growth is accelerating.
What many will hardly believe, but even third-quarter earnings are likely to show almost 30% growth. And year-to-date earnings per share revisions for 2027 forecasts have increased by more than 20% , while the S&P itself hasn't even added 20%. So the multiplier actually went down. If anyone thought the market was expensive at the beginning of the year , it actually became cheaper.
And, paradoxically, high oil prices, since the US is an exporter, actually strengthen the US economy. So I know it's a very complex situation, but I think there's a lot of pessimism because not only are the earnings forecasts good, but we're also seeing the leadership of stocks that are benefiting from lower inflation, like the tech sector and, you know, crypto.
You also seem ready for a market drop at any moment . It could still happen, and you've been talking about it for a long time . In fact, there was a point where you even allowed the S&P to drop to the 6,000 level before the end of the year and then come back up to the high 7,000s or 8,000s. So would that be some kind of midterm election panic or what could have caused that?
The midterm elections are indeed getting closer. I wonder how many of them there are? It was 35 days. Has another week passed? Is it 28? That's the name of a zombie movie. 20, 28 more days. Is it only about 28 days left? I mean, the correction has an expiration date, partly because the midterm elections are a denouement that is likely to be a powerful signal to risk investments regardless of the outcome.
And since June 30, I think the markets have been difficult. So, in part, I think we went through a gradual correction, and , you know, we had a lot of different groups that were in a state of decline. But as for a more serious downturn in the markets, I still believe that there could be some headwind from mid-October until the midterm elections.
Ugh. I'm not really inclined to make big statements because the sentiment has been negative and I think when people are pessimistic it's hard for markets to get ahead of themselves because we know that people actually act on their sentiment and remain cautious. Besides chips and data centers, what will be the next big thing that will fuel the tech boom in the fourth quarter?
Well, I think the main thing is starting to happen in the software segment . We are starting to see applications. This is why generative AI is so important. We saw winners during the "SaaS apocalypse"; the ones that survive are the companies that have, you know, basic transactional data. They have an advantage due to their enterprise software and sales.
They have the ability to renew contracts and can adapt generative AI in the workflow. So that's why Salesforce, ServiceNow, and Workday, you know, feel okay. They are returning to an upward trajectory. They see their stocks go up. But we're also seeing a number of new types of companies, called "AI-first " companies.
These are tiny teams. It could be 10-20 people bringing in 20 million in revenue. Some of them bring in 200 million in revenue. And that's the scale we're starting to see here in San Francisco. When I walk two blocks in one direction and two in the other, it's full of startups consisting of a few people.
They are called tiny teams, and they are capable of generating revenue on a scale we have not seen before. And that's exactly what I meant by the exponential scale that's happening now. These new "AI-first" companies operate very differently than everyone else. So you may not see 10,000 people hired, but in this market, 100 people can do the job of 10,000.
Ray, but this amazing surge in capital spending is being fueled by equity issuance. It is fueled by debt. We saw an almost complete loss of situational awareness. We have seen and continue to see this cyclical funding and the concerns surrounding it. We live in an environment of higher interest rates.
Oracle credit default swaps are at a record high. Will something break? That's the thing. I mean, companies that buy infrastructure and chips have very strong competitive advantages. These are digital giants. So, Amazon continues to generate free cash flow, and AWS is their most profitable market. So yes, they will continue to buy chips.
Look at Google in the advertising business, they are number one in advertising. We are talking about approximately 260 billion in advertising. So what's the point of spending a few hundred billion for them, right? It won't matter much to them . They will find somewhere to put it. They won't spend it on share buybacks.
And that's why only a few players will win in the AI market, and these companies are fighting for it. So this year we have 780 billion in liabilities. About 1 trillion will be spent next year, 1.3 trillion by 2028. And these numbers are actually very conservative. So, that's a huge capital expenditure, right?
I wonder how much of this is borrowed money or is under pressure from high global borrowing costs. Well, oddly enough, the corporate bond market is now competing with the Fed. Should I buy Google bonds or Fed bonds? Or perhaps buy bonds in Japanese yen? I honestly don't know. But at this point, corporations are creating serious competition for governments.
And it's actually a smart move—to raise financing, not necessarily use your own free cash flow. Oracle is probably an extreme case. You could say they are too over- credited. But they see it's worth making a big bet. And that's why only the biggest companies will be able to make such bets, and that's what makes the races so exciting.
Will this lead to collapse? This demand is real. It's not like the Internet, where demand wasn't created until the infrastructure was created. Here you can forecast demand for 3 years ahead , even 5 years. Let's say the 5- year figures are too optimistic. Even if you reduce them, you will start to see a payback in 2-3 years when companies start using the software and consuming the tokens.
Three headlines today, one topic. Everyone wants Nvidia chips. No one can get enough of them, and for some reason that still doesn't stop people from buying even more. The Indian company just ordered another 20,000 GPUs . Nvidia shares are heading for another record. And Katie Wood, well, she's actively buying up shares of one chip company and quietly getting rid of another.
Let's go. First, a company called AM Intelligence, part of the group behind Indian renewable energy company Grinko. Back in August, they placed a firm order for about 9,000 of Nvidia's next-generation Vera Rubin chips for a data center in Hyderabad. Now they've come back and ordered another 20,000 for India and Malaysia.
This brings the total to 29,000 GPUs and approximately 100 megawatts of power. If you're wondering what 100 megawatts means, that's a lot of electricity. This is a company that started with "green" energy and just realized: "Hey, we're really good at producing energy, and AI is really good at consuming it." And this is just the first step.
Their official plan is a gigawatt of computing power in India, the US, Finland and Malaysia, with over $8 billion for the first part. Their founder said, “Customers are essentially hunting for capacity. So when one company comes back for a second round a few weeks after the first, it's not just a trend.
It's a shopping cart that never stops filling up." Second, Nvidia shares are set for a new record, with headlines praising the key supplier as it celebrates the boom in AI demand. That's why it's important. Nvidia doesn't build everything itself. It's more like a chef who needs 100 suppliers to arrive on time.
So when a supplier says business is booming, it's like a farmer saying tomatoes are being snapped up. This shows that the restaurant is full before you even see the checks. Put this next to the first story and you get the same signal from both sides. Orders are coming in , and the people who produce the parts are busy working.
Third, the weekly Ark Invest review. Katie Wood bought more Nvidia and more Intellia, a gene-editing biotech company, and cut AMD's stake. If this looks familiar, it should . It has been selling AMD and buying Nvidia from time to time for many months now. So this is not an unexpected twist, it is a plot that continues.
Now, before anyone says that Katie sold AMD and AMD is doomed— slow down. Funds reduce positions for various reasons, such as rebalancing or taking profits. One week of trading is a blurry snapshot, not a verdict. So today the company is back for another 20,000 chips, Nvidia suppliers are celebrating, and one of the most famous fund managers is continuing to make the same bet.
This is also not financial advice. I make videos, I don't manage your money, and honestly, my own choices have had pretty mixed results. Whose side are you on, Nvidia or AMD? Write in the comments.
Comments 0