원문 보기
Good morning or good afternoon to everyone. My time zones are completely mixed up . I arrived this morning from Singapore. Yes, I got off the plane, landed, checked into the hotel, but they canceled my reservation because they thought I was supposed to arrive last night. So, if I look unkempt and disheveled, it's because of that.
I have a presentation. I'll keep a close eye on the time because I'll need to leave at 1:45 PM . But then I'll be back. So, I want to highlight some points. And the reason I want to go over these six points is because we know the reasons to be concerned right now: oil prices are high, bond yields are rising, and of course Ed has given his view on yields, and I, you know, agree with him .
But I believe that the market is a compressed spring. I think we will get through this, so I want to give you six reasons to stay optimistic. OK. The first is AI as a structural story. And space will be its continuation. You know, I know people are worried about AI, because, for example, this week there was this headline.
Firmus is one of the data centers funded by Nvidia. They are abandoning their IPO. And yesterday, again, I added something on the flight here because I wanted to add something. The FT published an article about OpenAI yesterday, alleging that the company was not meeting revenue targets, but that was untrue.
They simply misrepresented OpenAI's data. But, as you know, the media is not a friend of AI, they believe that AI is a bubble, but AI is doing amazing things. I don't know if you saw it. OpenAI announced this week that their model has solved 771 mathematical problems, representing 81% of all mathematical discoveries in the last 3 years.
And the average calculation time was 3 hours per task. It's like you said, "Hey, kid, you'll do your homework after dinner." And he took it and solved a bunch of math problems. And , of course, many mathematicians are upset. This is why people are afraid of AI, and I think it's justified. But for me, it's not a bubble.
The total costs being talked about are now 800 billion per year and moving towards a trillion. That sounds like a lot of money, but keep in mind: AI infrastructure worth $30 trillion is about the same as cellular and mobile networks, as well as road construction. I analyzed the wireless industry : handsets and networks cost 25 trillion.
If you add fiber optics and the Internet, it comes to over 30 trillion. So, AI equals cellular infrastructure . Commercial real estate—60 trillion. And China has spent 145 trillion in the last 25 years. Therefore, to me, 30 trillion seems like a perfectly acceptable figure. Of course, people are concerned about AI because of its impact on the job market.
I know you're all high net worth people here , but let me give you a little perspective on the job market and explain why I'm not too concerned about it. First, many studies, including one from Hubstaff, show that only 25% of the workweek, or 10 hours out of 40, is spent on actual work. Most people at work spend 3/4 of their time getting ready for work.
In addition, another study from Asana Group shows that only one in four jobs is real work. Most people work to help someone else do real work. That's it. So you don't have to worry about AI replacing jobs, because no one actually has a real job. We conducted research on frozen foods. Ken, our head of data science, did this work, and we found that when frozen foods came along , the share of agriculture in the total American workforce fell from 38% to 5% over two generations.
Think about it. 33% of jobs were replaced because frozen foods prevented food spoilage. And we haven't had any apocalypse in the labor market. In fact, it led to a huge innovation boom as farmers in the US were able to do other things. People are worried about AI because of debt issuance , and that's true.
We also know that the market sentiment is pessimistic. These are commercial short positions on 10- year US bonds. And this covers a period of 30 years. That is, this is a record level of short positions. I think that's why you shouldn't worry that the return on assets is incorrectly estimated. Many are pessimistic about returns, and if they are pessimistic, just remember that their pessimism will diminish over time.
And interest rates already have a significant impact on the economy. This is data from Jason Furman, an economist at Harvard, and it shows the various categories of construction spending in the US since the beginning of the year. You can see that data centers have grown by 64%. Everything else has decreased.
So, the Fed doesn't necessarily have to raise rates. This is already putting pressure on the economy. You have already raised the rates, the value of money. But data center costs are currently insensitive to higher rates. Therefore, I believe that raising rates is useless. But we are monitoring profitability.
OK. Now, for those who think this is a " bubble": I started working as a technical analyst in 1993, okay? And when I worked at Kidder Peabody, people first started talking about the public internet. And in fact it was 1994. So, you could see that from 1994 to 1997, Cisco stock went from 80 cents to $9.
And then they adjusted by 40%. And people thought this was the end of the Internet boom. By the way, it was a macroeconomic shock . It was the Asian financial crisis. But a year later, Cisco recovered to $18. But then the story with Long-Term Capital happened. Also Long-Term and default on Russian debt.
By that time, I was already working at Smith Barney. Salomon Brothers went bankrupt, and they merged with Smith Barney . So Cisco shares fell by half again. And people thought this was the end of the tech boom. But look what happened next. Cisco rose from $9 to $80 . That is, a 100-fold increase. Chat GPT was released in 2023, which means we are already 3 years into the AI cycle .
So, I think we are here. I... we're still at a very early stage. But I've been involved with mobile phones, so I can give you some perspective. This is the mobile phone industry. This is an S- shaped curve. You can see the growth from 56 million mobile phones when I started covering this to 9 billion.
That's 25 years of virtually exponential growth. But here's the thing. Wireless stocks peaked 10 years after the cycle began. That's the lesson . When we talk about AI, I think stocks will peak after 10 years of a cycle, even if it's 30 years of growth. And everything peaked at the same time. Operators, phone manufacturers, etc.
But after the bubble burst, it's worth owning stocks of lower-tier companies in the mobile or dot-com chain. Communication towers are one of the winners. You see that the towers fell along with everything else, but then they were far ahead of everything else. In fact, the towers are in the news today because Starlink is buying frequencies and wants to compete with mobile carriers.
And again , tower stocks are up again today. Another company lower down the chain is Apple. Apple didn't even get into mobile internet until 2007 . So, they missed the first 15 years of the mobile industry's development, but they had a nice interface and good fonts. They are really great for small devices.
In other words, as a story about the end product, their strengths only emerged as the industry matured. And of course, the best example of the end product was that, despite the fact that the internet was supposed to be the best thing, pizza—Domino's Pizza— outperformed all other promotions. Um, I'm optimistic about the earnings.
I'll cover this quickly because I only have 5 minutes left before I have to go. But, uh, the profits are good. At the beginning of this year, the expected earnings for 2027 were 350. And now the earnings for 2027 are 415, but by the end of the year we will probably be at 435. So from the beginning of this year to the end of the year, next year's earnings have increased by 25%, and the market has not increased that much.
This means that the market has become cheaper, you know? Um, I'm going to show it. Uh, and here's the thing. The market has faced all these challenges over the past 6 years. COVID, the effects of supply chain disruption, the fastest inflation cycle in history, the fastest Fed rate hike in history, last year's tariff wars, last year's US bombing of Iran's nuclear facilities, and of course, this year's US-Iran war .
All of this should have been a threat to profits and the market, but the market held out. And usually, if a company goes through so many challenges and survives, you have to say it's less cyclical, so the multiplier should go up. The S&P's P/E is actually lower today, despite having gone through all these tests.
This is a reason for optimism. OK. Um, the Fed is maximally hawkish, which means maximally painful, and that's a positive. I'll skip this because, um, the Fed is showing determination to raise rates. Um, but, you know, I really think that ultimately the Fed is under pressure, uh, economists are putting pressure on the Fed right now.
Um, because 50 basis points won't kill the economy or the stock market. But I want to move on to talking about the midterm elections. Uh, and there's a certain dynamic here, because I think it's probably fair to expect that the Democrats will take control of the Senate and the House of Representatives .
But for AI stocks, it's actually not bad. Let me explain. So, these midterm elections will be held on November 3rd. And the market is predicting that, um, the Democrats will take the Senate and the House of Representatives. And , of course, there are reasons: oil, the K-shaped economy, and data centers.
But here's the thing. Today, Republicans support data centers, and they supported the Clarity Act. Democrats opposed both. It was beneficial for Democrats to oppose data centers because it resonated with voters. But it is also well known that China is behind much of this backlash against data centers.
And it is also well known that Democrats supported the Transparency Act. They simply didn't want the Republicans to claim victory. So, after the midterm elections, I think Democrats will support data centers and the Transparency Act . So, you will see a change in political course. I believe this will contribute to growth in the field of AI.
And, of course, the overall market should go up. This is what the markets look like in a normal year. This is what they look like in midterm election years. But I'll skip ahead : they bottomed out in early October. However , with Republicans controlling the House and Senate at the beginning of the year , as well as the White House, markets bottomed out in early November.
This is what Mark Newton and I are sticking to : we believe the market can go up until mid- November and then take a break. And, well, let's see. Okay, in a few minutes I'll cover the last part, which is: we expected this year to be a three-phase market, and that's exactly what's happening. We thought the market would look like this at the beginning of the year, that we would go up and maybe get to around 8,000.
We thought about 7,300, but the market had enough strength, so we think it's around 8,000. And we think there are reasons to pull back later, because the market is always testing the new Fed. We had a war going on. We also have major IPOs and share unlockings coming up. We have a huge amount of margin debt, but there was already a massive margin call on June 30th as Leopold's fund received a huge margin liquidation.
But I think profitability can certainly reach your breaking point. So I think all of this is creating a bear market feel. But I realize that a bear market of sorts has already occurred in the Magnificent Seven, cryptocurrencies, software, and AI. And then we'll be steadily rising through the end of the year, approaching 8,000.
But part of me thinks that maybe most of that correction has already happened, because the top chart shows the S&P 500 over the last 6 months, and you can see it's looking great. I mean, it's almost at historic highs. The bottom chart is the S&P 500 with equal shares. So, this effectively eliminates the concentration of market capitalization.
And here you see that the equal shares index has already fallen 8% from its highs. So, essentially, I feel like the bear market is already here for many groups . Good. Um, and one last thing: I think rich people don't invest enough in stocks. This will be the last thing I will talk about. Okay, and I'll be back soon.
I just need to, um, do an interview with CNBC. That's $200 trillion in wealth in public stocks. This is data from State Street. I want to put this number in perspective. If the net worth of households in the world today is 200 trillion, and these are just liquid assets, because of course they own businesses, art, and other illiquid things.
Annual investment income at a rate of 5% brings in 10 trillion per year. So, in other words, this number is growing by 10 trillion every year. So if data centers consume 800 billion, and the state deficit is 2 trillion, there is still about 7 trillion left, which is constantly looking for somewhere to invest.
So if you're wondering why things are going so well, that's part of the answer. So, I think the money comes from private equity. There, people are over-leveraged, and under-leveraged in public stocks, because in the public world, half of all liquid assets are stocks. If it's not 50% of your allocation, then you're underinvested in them.
And one more thing, the last thing I'll show is that the transfer of wealth plays a big role. Vlad Tenev says that a $100 trillion wealth transfer is currently taking place . And Ken also estimated that it was about a hundred trillion. But the thing to remember when transferring wealth is that approximately 65% of all wealth is controlled by people over the age of 65.
And in the next 20 years, they will mostly be inherited by millennials . The U.S. Internal Revenue Service actually shows that this pace is accelerating. And you can see that it's really accelerated over the last 3 years , and I think that's supporting the stock. And the last graph I want to show as a reason for our structural optimism: the US is benefiting from a growing number of working-age workers.
That is, people aged 30 to 50. This data dates back to 1935. You see that the number of people between the ages of 30 and 50 is quite cyclical. The minimum of this indicator occurred twice. Once in 1967 and the second time in 2009. Good. This has a high correlation with stock returns. You can see that the 10-year rolling stock return matches this exactly.
This is because people between the ages of 30 and 50 are the most productive in society. They get married, get promoted , have children, and drive economic growth . This graph shows that if you overlay each generation and population peak, it coincides with each bull market peak. So the peak of 1929 coincided with the peak of the " Greatest Generation." The peak of the " silent generation " fell on 1974.
The peak of the baby boomers was in 1999. The peak of Generation X was in 2018. In fact after 2018, you know, we were actually trading in a range until 2022. And the next peak will not be before 2038. So, again , that's why we're structurally set up for growth for many years. And as for goals, this is our schedule, which we first published almost 10 years ago.
In 2017, and our view is that by 2029 the S&P could be somewhere between 10,000 and 19,000. So if we take the midpoint of 15,000, remember that 10 years ago the S&P seemed to be around 1,600 or so . Yes, he was much shorter. People thought we were crazy because we were talking about 10x growth, but as you know, we're almost at 8,000.
It's a really big leap to go from 8,000 to 15,000 in the next 5 years. So, that's why we remain structurally geared towards growth.
Comments 0