저자: KatieWood · 외국인
요약 보기
AI와 기술혁신으로 미국 명목 GDP 성장률이 6~8%(중기 7%+)까지 오를 것이라 주장. 생산성 향상으로 인플레이션은 장기 하락, 유가는 전기차 확산으로 하락 전망. 수익률곡선 역전은 일시적이며 장단기 금리는 높은 명목성장을 반영해 상승 가능하다고 봄. 엔비디아·브로드컴 등 AI 수요가 공급을 계속 앞설 것으로 보며 마이크론 실적 긍정 평가. 반면 마이클 베리는 엔비디아·마이크론·팔란티어에 풋옵션을 걸어 단기 급락을 경고하고, 엔비디아의 자체 자금조달 구조에 대한 우려도 제기.
다듬은 글
Kathy Wood of Ark Invest wrote a letter to investors earlier this week with many interesting thoughts. Let's congratulate Katie Wood. I will not read anyone's story. I want to go straight to you . Katie Wood joins us now for an exclusive interview. Katie, I'll tell you how I summarized your words. The US economy has weathered rising interest rates, an inverted yield curve, tariffs, a government shutdown, and war.
It continues to show high growth rates, and you expect the same in the coming years with an additional factor that could rival the industrial revolution—artificial intelligence. Absolutely. In fact, we are building on the experience of the Industrial Revolution. If you look back 400 years before the Industrial Revolution, Bret Winton, our chief futurist, did this kind of work.
Real world GDP growth was about 0.6%. And then, with the beginning of the industrial revolution, which was actually the last technological revolution, we reached 3% of global GDP growth. So that was a fivefold increase. We believe that over the next 3-5 years we will see regular growth, averaging over 7%.
This is only a two-and- a-half-fold increase, despite the fact that much more technology is being developed today than then. Back then, it was electricity, the telephone, and the industrial internal combustion engine. Now we have artificial intelligence, robotics, energy storage, blockchain technology, and multi-omic sequencing in the life sciences.
So, a lot more technologies are moving a lot faster. So I actually think 7% will look conservative in the long run. We have a lot of other news, and here we have the straits closed, and, you know, fears about inflation are on everyone's lips. You believe it is on the verge of a long-term decline. Why?
Yes, I believe that productivity growth, in particular, is one of the reasons. We also believe that oil prices will fall significantly as transportation shifts to grid-powered transportation . And the network does not use oil. 57 % of oil consumption is related to transport. So we believe that as transportation transitions to electric grids, demand for oil will fall, and we have a chart that shows that during a crisis, the price of oil typically drops to $30-35.
We think this won't be a crisis, except perhaps for oil companies and energy-dependent countries , but we think it could happen again , but this time it will become a sustainable trend. During the Industrial Revolution, we had an inverted yield curve that didn't cause a recession, and you expect something like that again.
Can you explain why short-term rates are rising while long-term rates are expected to remain stable? Why is this happening in the economy that you predict is doing this? So, again, taking lessons from the Industrial Revolution, short rates—that was for the 50 years ending in 1929 , until the Depression.
Short-term rates averaged nominal GDP growth, which was then 4.8%. Long rates averaged 3.8%, so short rates were oriented towards nominal growth. Long rates then depended more on real deflation or the absence of inflation. So the yield curve was inverted over 60% of the time by an average of about 100 basis points.
Therefore, we believe that we can see nominal GDP growth in the range of 6–8%. This is the level we believe short-term rates will reach, while long-term rates will be at 5–7%. So, from a government bond perspective, we are already at the lower end of that range. I don't know if you saw Joe Lonsdale the other day, I said we're talking about $40 trillion in debt, and it's such a daunting number that it seems like there's no way out.
He said not to worry, you know, everyone wants to raise taxes. Don't raise taxes. We will enter a disinflationary decade starting in 2030 thanks to AI. Growth will be so high that you won't even have to raise taxes, and everything will take care of itself. Did you see how he said that? Katie, do you agree?
I didn't see him say it, but I generally agree with the idea that we will get out of this situation through growth. I was at the beginning of my career during the Reaganomics era, and the same controversies were going on back then. And, of course, growth solved the problem then, and I believe it will solve it again.
You know, it's interesting, I think many people will start competing for this growth rate. We are in the 7–8% range, and I saw Elon Musk, I think he is predicting over 20% real GDP growth. So I agree that the surprise will be how fast this growth will be. And we're already seeing this from AI companies.
They have amazing growth rates. Micron's reports just came out, shares are currently up slightly after the company beat revenue and profit expectations. The company reports that its revenue and earnings per share guidance for the next quarter is better than expected, but gross margin in the first quarter may be lower than forecast.
I'm here with Matt Bryson from Wedbush. So, Matt, the bar was pretty high, because even with the volatility in stocks in recent months, earnings estimates and expectations were strong. What is the most important thing in this report? I would pay attention to the earnings forecast. They expect over $38.
I think this is roughly what investors were expecting. I've heard predictions in the $ 36 to $40 range. So we are right in the middle of that range. I think this is the most important indicator from my point of view. There is only a slight decrease in expected gross margin for the current quarter. I think this is a question that will continue to keep investors on edge about how long the price advantage, supply constraints, and demand will last.
Yes, it really all depends on the details. For example, if this reduction is due to the supply of more high- bandwidth memory. That is, AI support. I don't think this is a negative factor. All my price checks indicate that prices are still rising . Although at a slightly slower pace . So I believe that the answer to the question "why" is crucial here.
Of course. And, you know, as earnings expectations continue to rise , stocks look cheaper than they have been recently. The question arises: are we approaching a point where the market will start to worry about capacity expansion in the coming years and will we see some slowdown in data center development and macroeconomic demand drivers?
Yes, I think there are certainly concerns about new capacity coming to market. Supply is always cyclical. Only at the end of 2027-28 will we see these capacities begin to fully operate . However, it is important to remember that we are getting huge predictions about the creation of accelerators and processors.
So when you think about Broadcom, for example, where demand doubles and then doubles again, and the amount of memory for each of these compute chips or accelerators stays the same, the increase in bits is not even close . So as long as you have that much AI power, even if supply increases, it's not clear to me at all whether it will catch up with demand.
And in terms of capital return, I know there may have been expectations that Micron would update information on how much of that profit that it's generating in free cash flow it's going to distribute. Yes, I think the best way to return capital to shareholders would be buybacks, but they are blocked until December 9th because of the CHIPS Act funding they received.
So I'm not sure we'll hear much new on this call, but I expect they'll buy back a lot of shares, given how much cash they've accumulated, once we get past the December 9th date . Apple, Alphabet and Nvidia are leading a broad rally in tech stocks today as interest rate concerns ease. When expectations for rate cuts strengthen or at least stop deteriorating, money typically flows back into stocks, which are most sensitive to the cost of borrowing and future earnings.
And it is to this group that Nvidia belongs. It wasn't news from one company that caused such a movement. This entire sector got the green light at the same time. Second, CoreWeave shares are rising on news of expansion with Nvidia's Vera Rubin. CoreWeave has built its reputation on being the first to adopt Nvidia's latest hardware, and Vera Rubin is the next- generation platform after Blackwell , combining the new Rubin GPUs with Vera CPUs for a serious leap in inference performance.
Nvidia also owns an equity stake in CoreWeave and is directly investing in the development of its data centers. So every such expansion benefits both companies simultaneously. CoreWeave is the first to receive the latest technologies . Nvidia is getting a flagship customer that will test its next platform in real-world conditions.
And now, the other side of today. Michael Berry, an investor known for predicting the 2008 housing market crash, said he bought put options on Nvidia, Micron and Palantir, essentially betting that all three stocks would fall . He warns that the market's rise to record highs could trigger a 1987-style crash, where stocks fall sharply and quickly, rather than slowly sliding downward.
What is notable is the change in how he places bets. Berry recently said he converted his short stock positions into put options , explaining that this allows him to increase leverage over a shorter time horizon. This is a sign that he believes any downturn will happen sooner rather than later. He also openly admitted that this agreement has not yet yielded results.
By his own account, every one of his AI- related short positions has been profitable, except for one—against Nvidia, which continues to grow against his expectations. Here's how you can combine all three stories at the same time. The stock's rise and the news about Coreweave are all about real current demand, money, and equipment moving right now.
Berry's bet is about cost and timing. He's not saying the technology is fake, he's saying the price has outpaced the fundamentals, and that the correction could be sudden, not gradual. So, today Nvidia shares rose, Coreweave expanded its fleet of flagship hardware, and one of Wall Street's most famous "bears" bet against all of this at the same time.
Both statements may be true: now there are strong positions, but at this price there is vulnerability. Let me know whose side you are on. Is this boom as strong as the headlines suggest, or is it largely driven by Nvidia's own money and marketing ? Let's figure it out. First , a new report reveals the true cost of Nvidia's AI Agent Security Platform —a system they launched last week to combat unmanaged agents.
The catch is that this protection runs on specialized Nvidia hardware, the BlueField DPU, so companies that need this security have to buy more Nvidia chips. So the proposition is not just safer AI agents, but safer agents with increased reliance on our stack. This is a smart business move. Prices haven't even been announced yet, but that means the same company that sells you computing power is also selling you a hedge against the risks that power creates.
Second, and more importantly, Nvidia's plan to fund the entire AI boom is facing resistance. Back in August, Nvidia agreed to more than $500 billion with Wall Street giants like Blackstone, Apollo, and BlackRock to allow AI companies to borrow against chips, treating GPUs like real estate or toll roads— assets that retain value for years.
Lenders don't fully believe this. Banks typically amortize GPUs over 3–4 years, while Nvidia claims its chips can generate revenue for up to ten years . This gap is important because it determines how much debt can be raised for this equipment, and some lenders are now requiring stricter guarantees and contracts from customers before making deals.
Analysts also pointed to a larger problem. Nvidia's financing of its own chip purchases by customers risks inflating real demand compared to what exists only because Nvidia sponsors it. Third, Nvidia, AMD, and Intel rose today after Micron's report, but Micron's stock itself fell. This is a pattern that has occurred in this sector before.
The memory supplier's strong results confirm that AI demand for high-bandwidth chips and memory is real, supporting chipmakers who are benefiting from this demand. However, Micron's own stock often falls due to profit-taking immediately after a strong earnings report, especially after significant growth, as investors sell on the news rather than reward it.
So, today Nvidia is selling security patches for a risk it created. Wall Street is becoming more picky about whether to trust Nvidia's own chip cost figures. And the market continues to reward the ecosystem around Nvidia more than the company that supplies the memory itself . None of this ruins the story around AI, but it's exactly the type of friction worth keeping an eye on. Let me know what impressed you the most.
Comments 0