저자: TomLee · 외국인
요약 보기
Tom Lee은 미국 기업 이익 개선(1분기 S&P 주당이익 $10 상회, 연 $40)과 AI 투자 사이클로 S&P 800p 상승 여력을 주장하며, 2026년 말 조정 후 중간선거 이후 2년간 생애 최대 주식 랠리를 예상함. 반도체·제조업 PMI 상승은 AI 자본지출이 견인하나 소비심리 약화와 실질소득 감소는 리스크로 지적. 이후 Schwab의 Chris Ferraro가 제조업 회복과 AI 자본지출 사이클의 지속을 설명함. 후반부는 엔비디아 재평가 기회와 마이크론의 구조적 변화(HBM·데이터센터 메모리) 가능성을 언급.
다듬은 글
I am speaking. I don't need notes to talk about you because you show up often and I remember what you say just to set the stage for what we're going to talk about. In the past, you have made predictions that people rarely dare to make. For example, that the inflation rate will be good on Friday. Or that the S&P will rise 8% this week.
And many times this has come true, and you have a reputation for being a fearless person in this regard. Here's the last prediction you made, and it came true, although it does n't show up very well on that chart, but you were optimistic for the year , and things were going pretty well until the war started, and I remember it was almost like a mini-collapse on the day of liberation.
Not that bad, but pretty steep, and you said that's it , and by the end of this month we'll be much higher. And this was exactly the moment when everything went wrong. Not only did it happen, we took off in April, as you said, and continued into May. So, I mean, you can have a real circle of honor, and we have so many people coming in who have no idea what the market is going to do in the short term or the long term, and it just contrasts with the fact that you do so well from time to time.
Although you can't do this forever , Tom. One day you will be wrong, but this time you will be right again. Yeah, well, thanks, Joe. I know it was a tough market for people. I'm not surprised. We expected 2026 to be a year where turbulence is possible and people will encounter misleading signals . But there are still very powerful favorable factors for America.
One of them, of course, is the story of artificial intelligence. And now we're learning about energy independence, right? America can cope with high oil prices because we will not have a shortage. And then there are new positive factors emerging as AI spreads further, it really benefits American businesses.
I think that's why S&P earnings were $10 higher in the first quarter alone. That's $40 per year. And if you simply apply a factor of 20, it will add 800 points to the S&P. So I think there are many reasons why stocks could recover. Um, well, yes , but then again , I think we could still have difficulties between now and December.
So I think everyone should still be vigilant, but optimistic. What did you say about the next two or two and a half years being some of the best market performance we've ever seen ? Do you still think so? And that was, that wasn't in this program. You said that elsewhere. But what, what, what is your forecast for the next two years?
And what is it based on? Well, I think, Joe, there are a few factors that are coming together that will really support once-in-a- lifetime things. First, I believe that the US economic growth rate is actually starting to pick up. You know, in other words, we could be growing at 4%, and for the mature largest economy in the world to start accelerating growth—that's pretty impressive.
Second, the US is one of the largest exporters of the most important tool of the next 10-15 years, such as AI products, which means that we are effectively a net exporter of a high-value product. And today, in my opinion, so much capital is misallocated, with a significant portion of it being held in private alternatives, but it will move into the public market.
So I think that, combined with the demographic push of millennials and Gen Z entering the workforce and also starting to inherit wealth , I think that will lead to the fact that after 2026, maybe in the next 2 years, we'll see the biggest stock market rally of our lifetimes. We just had a board. So, you think we'll hit 7,700 in the near future, and then there's a pullback, and after the midterms we...Does this coincide with the midterms, this pullback, and then we go up?
Is this, is this correct what you are saying? Yeah, so that's, uh, our best guess. You know, I think the markets could still add to their gains in June, just not at the same pace as they did in May, and I think that's... The domestic growth is largely driven by Korean memory stocks, Samsung and SK Hynix.
For more information, we invite Chris Ferraro, Head of Equity Research and Strategy at the Schwab Center for Financial Research. Chris , we are glad to welcome you here. So global growth seems to be accelerating everywhere. What is driving this improvement? Well, this is an improvement—a change from what we've seen in recent years, because now we're seeing a real upswing in the strength of the manufacturing sector.
And this is not just about capital investment in artificial intelligence, we are seeing how the growth in production is truly spreading across most regions and industries. Chris, if we look at this shift from service-based growth to manufacturing-based growth, why is that important? What are the implications for the global distribution of assets of expansion driven by manufacturing compared to that driven by services?
Yes, that's a great question. I think there are several key reasons. First, if you look at the composition of global stock indices relative to the economy, they typically have a higher proportion of manufacturing sector companies, while the US economy, for example, has become much more service-oriented.
Financial markets, global equity markets, have significant exposure to the manufacturing sector. This applies to sectors such as energy and industry. Of course, so does the technology sector , given the AI capital investment cycle. Also basic materials. You can look at the list and see that the strength of manufacturing is reflected in corporate profits.
And to what extent is this revival of manufacturing directly related to spending on AI infrastructure? Certainly, to a large extent. We saw this in the semiconductor earnings spike , as you noted, semiconductor stocks in emerging markets and the U.S. are showing triple-digit earnings growth this year due to the shortfalls caused by AI spending.
But it goes far beyond that . We are seeing strong earnings growth across a wide range of industries with this exposure. This applies to classic manufacturing sectors such as industry, energy, raw material suppliers, basic materials, telecommunications, and, obviously, utilities, which are also connected to AI through the development of capital infrastructure for AI data centers.
So we're seeing quite a broad force, but the concentration of gains—the real numbers of strong growth that we're seeing—are definitely tied to the AI capital investment cycle. Chris, the graph we see on the screen shows that earnings expectations have risen sharply. Which regions currently have the strongest forecasts ?
Yes, this profit boom is not just about expectations. This, this is also in actual earnings, if you look at the period over the last 12 or 24 months, we are seeing some of the strongest earnings growth, especially in a mid-cycle dynamic, that we haven't seen in a long time. In addition to this, there are very high expectations associated with capital expenditures for artificial intelligence.
And why have these emerging markets benefited so significantly from this AI boom ? And, perhaps, my second question: is the performance of international stocks now becoming increasingly dependent on demand related to artificial intelligence? Yes, I think there are a few points worth noting here. The first is that the capital expenditures of hyperscalers investing in data centers and seeking to develop artificial intelligence are causing a shortage of semiconductors.
And this shortage is pushing semiconductor prices to cyclical highs . And this stimulates profits in these sectors. But, you know, in the chart that you showed earlier, a significant contribution to earnings growth, expected earnings growth, is coming from the technology sector and just a few semiconductor companies.
But you also see that most of the other bars on this graph are positive, which means there is a broad range in terms of earnings growth. It's just the scale of what we're seeing in terms of growth in the semiconductor industry trumps almost everything else. And when we think about concentrating on global equity markets, there are several ways to approach it.
First, in terms of the absolute size of certain sectors now relative to the broader market. So if you look at the S&P 500, for example, or emerging markets , 40-50% of these indices are now made up of technology and communication services sectors, which is significantly more than it has been historically.
And then there's the concentration of profits, and that concentration that's really driving the growth of profits here comes from those same sectors. We've talked a lot about growth, but I want to talk about the disconnect we're seeing between growth and consumer sentiment . Because if we look at the global purchasing managers' indices (PMIs) in manufacturing, they are rising.
But if we look at global consumer sentiment, it is noticeably deteriorating . So, how long can this manufacturing boom last while the end consumer feels the economic pressure? Yeah, I think, you know, there are some limits to how long this capital investment cycle can last. We see hundreds of billions of dollars being spent by just a few hyperscalers.
And consumer sentiment has been weak for some time now , and of course that hasn't had much of an impact on earnings growth, but it's a combination of those factors. First, real incomes stagnated. Although the labor market is currently quite stable, if you look at incomes adjusted for inflation, they have started to fall into negative territory.
This is due to rising resource prices and inflation. When you survey consumers about food prices , gasoline, housing, it definitely becomes a pain point. Chris, you mentioned hyperscalers and brought up a very interesting statistic: the capital expenditures of just the " big five" —Oracle, Meta, Microsoft, Alphabet, and Amazon—are approaching 2.5% of US GDP.
If they decide to optimize or slow down their investment spending in late 2026, how will that impact the global production expansion that you are currently tracking? Yes, that would definitely slow him down, no doubt about it. And I think we can say pretty confidently that these costs don't look like they're going away anytime soon.
In fact, if we look at 2027, there is a planned increase in these capital expenditures. That doesn't mean it doesn't pose a risk to the markets. That's obviously true given the heightened expectations, but we don't anticipate, you know, any sudden collapse in the near term related to this cycle. Chris.
Okay, I hope you're doing well today and keeping calm in this market. Today we need to talk about perhaps one of the most important rotations taking place within the artificial intelligence trade. Because while memory stocks are experiencing tremendous volatility, something ver. It looks like the money is flowing back to the biggest companies in artificial intelligence.
Nvidia is waking up, Broadcom is attracting buyers, and the "Magnificent Seven" is starting to show better results. And at the same time, investors are debating whether the extraordinary rally in memory stocks such as Micron and SK Hynix still has the potential to continue. So, today we will examine both sides of this debate.
I believe the next few months will show whether the AI infrastructure boom is entering another growth phase, or whether some of the biggest leaders are finally approaching a point where investors should be more cautious. Let's start with Nvidia. In recent months, one of the main sources of uncertainty for the AI sector has been the capital expenditure of hyperscalers.
Amazon, Microsoft, Google, Meta—all these companies are spending huge amounts of money on building data centers and purchasing infrastructure for artificial intelligence. And investors keep asking the same question. When do expenses become excessive? When will shareholders finally say enough is enough?
And when will these huge capital investments start to hurt the stocks of the very companies making them? Meta was perhaps the most glaring example of this dilemma. Every time Meta increased its spending plans, investors got nervous. Stocks were punished by the market as participants doubted whether companies would be able to get sufficient returns from the hundreds of billions of dollars invested in AI infrastructure.
But something has changed. Fears about capital spending appear to be easing. Investors are starting to believe more that hyperscalers can generate attractive returns from their investments in artificial intelligence. And if the upcoming reporting season confirms that cloud revenue is growing thanks to demand for AI, it could be a hugely important catalyst for the entire tech sector.
Because the AI investment cycle works like a chain. New AI models require more computing power, and the need for power requires more GPUs . More GPUs require more memory, and more memory requires additional semiconductor manufacturing capacity. Larger semiconductor production requires additional equipment.
More data centers require more energy, and more energy requires additional infrastructure. And as long as companies spend hundreds of billions of dollars on infrastructure, proving that it generates attractive profits, the entire chain can continue to grow. This is why future earnings reports from hyperscalers are so important.
If Amazon, Microsoft, Google, and Meta demonstrate accelerating cloud growth and improving returns from AI infrastructure, investors may become more confident about their massive spending plans, and that could drive capital back into the biggest tech companies. We are already starting to see the first signs of this.
Apple is hitting new highs , and Broadcom is attracting buyers. And Nvidia is starting to "wake up" after showing almost nothing for several months. This is what makes Nvidia so interesting right now. The shares are being consolidated. They are holding near their 200-day moving average. The valuation has declined, and compared to many other semiconductor stocks, Nvidia is showing significant underperformance.
This does not mean that the company has stopped growing. Quite the opposite. Demand for Blackwell remains extremely high. The company continues to expand its infrastructure platform for artificial intelligence. And Nvidia is nearing a point where quarterly revenue could eventually reach $100 billion.
Think about how incredible that is. Nvidia has transformed from a company known primarily for gaming GPUs to a major infrastructure provider for the AI economy . And yet, after the huge rise in other semiconductor stocks , Nvidia has become one of the most controversial outsiders in the sector. Some investors see this as a warning, others as an opportunity.
And I think that's how investors should perceive it. Nvidia and Broadcom are almost becoming anti-impulse trading in artificial intelligence. Investors who had chased the growth of these companies for years have now begun to reassess the former leaders, now that shares of memory and semiconductor equipment manufacturers have delivered huge profits .
And historically, buying quality companies after long periods of consolidation can create interesting opportunities. But Nvidia is n't the only stock investors are watching. Micron remains at the center of the AI memory trade. And the discussion around this company has become extremely interesting.
Micron recently delivered one of the strongest earnings reports we've seen during the AI infrastructure cycle. The new growth was extraordinary. Margins increased significantly and demand remained high. And management continued to emphasize significant supply constraints. The company has actually sold off most of its production capacity.
Customers sign long-term contracts, and AI infrastructure continues to consume huge amounts of memory. That's why some investors believe Micron is no longer just a cyclical semiconductor company. They believe that the company is undergoing a structural transformation. Historically, Micron has been heavily dependent on the consumer market, personal computers, smartphones, consumer electronics, and traditional storage devices.
And these markets were cyclical. Demand was growing, supply was shrinking, and memory prices were increasing. Producers expanded capacity, supply eventually exceeded demand, and prices collapsed. Margins were disappearing and stocks were plummeting. This is a traditional memory cycle. But artificial intelligence can change this equation.
Micron is increasingly focusing on higher-end memory products, high- bandwidth memory, data center memory, artificial intelligence infrastructure, and automotive memory.
Comments 0