저자: TheLongInvestor · 외국인
요약 보기
AI 섹터 급락 속 CoreWeave는 부채·금리 부담으로 하락했으나 계약잔고 1,040억달러, 매출·조정EBITDA 성장과 이자비용 대비 커버리지(2.4배 이상)로 펀더멘털이 탄탄하다고 주장. 젠슨 황의 투자(약 20억달러)와 캐시 우드의 1,350만달러 매수를 근거로 저가 매수 기회라고 봄. Gartner의 AI 투자 회의론은 과거 예측 오류를 들어 반박. VIX 25 이상 등 바닥 신호를 언급하며 공포 구간의 매수를 제시.
다듬은 글
The stock market is falling. AI stocks are plummeting, and I'm going to find out why by taking apart this little company called Coreweave, okay? Because they are at the very center of it. I'll break down the chart for you, tell you what research firms and some of the super investors who are buying up stocks on this dip right now are saying, why they're doing it, and what CoreWeave's CEO is actually saying about his own company.
We will also hear from Jensen Huang, who is essentially the architect of this entire AI boom. So, as you can see, the Coreweave graph is incomplete, you see? The reason is that it was one of the most popular IPOs during the AI boom, so to speak. Back in May or March 2025, Coreweave went public with a price of around $60 per share.
It was an extremely successful IPO. Eventually, the price jumped to the first wave, somewhere between $195 and $200 per share, before gradually fading away. As you can see, we never really broke out and started a new impulse wave until our third wave, which, as you know, after reaching the second wave—and we expect around $45 per share—will be the most aggressive impulse move.
As you can see, we bounced back one, two, three, four, five , let's say six, maybe seven times during these aggressive sector-wide sell-offs that have been haunting AI for the last year or a year and a half. This is not unusual, but every time it happens, it seems like the world is falling apart, the market is crashing , and everyone is panicking.
And, unfortunately, many people lose money in such situations, but this is historically a very attractive scenario in terms of risk-return ratio. What is the reason for all this happening? It's all about this graph, you know? Coreweave has a huge pile of debt on its balance sheets. I think it's almost 30-45 billion.
I'll break down the numbers in a moment, but their borrowing costs have skyrocketed . So the market is saying, "Hey, Coreweave, we need more insurance on your debt because we don't believe you'll remain solvent." And, as you can see here, back in early September their bonds were at about 12–12.5 %, and now they've just skyrocketed.
Now, one of the main things we can point to is this little one, which is the 10-year US Treasury note , okay? As 10-year bonds, as I've pointed out in many videos, continue to rise , it puts pressure on companies with large debt loads. We saw the 10-year bond peak around 5.31 due to US inflation reports.
And this, as we know , was caused by the war in Iran. Well , now we are starting to see how this is really affecting, the market is starting to demand certain guarantees, and that is exactly what happened. So we see that companies that are so closely tied to AI are forcing investors to demand higher premiums on these returns.
And when bond yields go up, you get a downward inverse correlation, and that's why CoreWeave fell. At what stage are we now in this AI trade? And there's research firm Gartner, which seems to have added fuel to the fire as to why this is happening. Over $3.5 trillion in AI infrastructure spending could be at risk of being devalued by the end of 2030.
So they stated that by 2030, 80% of new cloud providers are expected to cease independent operations. They also noted that over $3.5 trillion in AI infrastructure spending would be at risk, as I said. They said: “In 2026, AI vendors are expected to spend nearly $1 trillion on AI infrastructure, and are projected to invest an additional $6 trillion between 2027 and 2030.
” And a senior analyst at Gartner argued that their research indicates that revenue from AI for enterprises and consumers will be insufficient to provide a profitable return on all planned capital expenditures on AI. They said: “ By 2030, AI data center owners will need to generate a combined $8.5 trillion in revenue to cover announced capital expenditures, financing costs, and a 12% return on invested capital .” So they're saying that this AI development is going to cost a lot more than, you know, the return that investors will get .
The only way we can see this coming back now is to look at the profits of hyperscalers. We see the growth of their AWS at Amazon. We see Google Cloud growing, well, up to 80, 81%. We're starting to see Meta using new, more live products like Meta Muse and their AI agent, and it's just dominating the headlines about how everyone is using Meta Muse and, you know, how much more efficient it's made their lives .
Meta's investment in AI is starting to pay off, you know, not just through advertising, but now we're seeing real products emerge. And this is like the next wave of AI development— consumer-oriented products. So, with that in mind, I found a funny tweet here. Congratulations RJC Capital. They essentially analyzed whether he analyzed Gartner's track record and some of their " greatest hits." They predicted that by 2011, 80% of Internet users would have an avatar in Second Life.
It was zero. That was zero, sorry. They predicted that by 2015 BlackBerry would have 11.1% of the market. The result was less than 0.5%. They predicted that by 2018, 3D printing would cost the world at least a hundred billion dollars a year through intellectual property theft, in reality closer to zero.
It was close to zero, sorry . They predicted that by 2020, customers will manage 80.5% of their relationships with companies without human assistance. By 2018, they predicted that only 25% of customer support services would use virtual assistants. Okay, we see that this is not true . And they predicted that by 2026, 25% of people would spend an hour a day in the metaverse.
I have n't seen anyone in the metaverse at all, and, you know, it's obvious that Meta has closed that avenue. These are just some of the, you know, I'm not criticizing Gartner, but they like to make these kind of outlandish predictions, and obviously not all of them turned out to be very authoritative, to put it mildly .
That's the biggest concern, and as the headlines start pouring in, we start to see it, you know, investing.com, Investopedia, all these names, CNBC, just churning out these articles, publishing them all at the same time, and that puts additional pressure on stocks. So, who came forward to defend? The CEO of CoreWeave publicly stated that this news is completely irrational, and, you know, the stock apparently fell 17% on Monday.
At the time of recording this video on Friday, they had fallen another 12.5%. So, 29% in total, they closed yesterday, but the CEO essentially said: when you build a new type of technology company, there will be periods when the market doesn't fully understand everything and acts irrationally. The indicator on the screen does not always reflect the real state of affairs of the company, and if you are a long-term investor like us, this can be a very good opportunity.
The next person to speak is Nvidia and Jensen Huang, you know , the CEO of Nvidia. Keep in mind that Nvidia has some shared interests here. They recently invested nearly $2 billion at a price of about $87 per share. What 's more, you can now get in at a lower price than Nvidia. So now, if you bought CoreWeave today , you would get a 33.9% discount compared to Jensen Huang and Nvidia itself.
I think that's a good risk-reward ratio that I'm willing to take, okay? Jensen is one of the strongest minds on the planet. I think we have to take this extremely seriously . What did Jensen actually say? He came out and declared, "The strength of CoreWeave is growing every day." He said, "I'm optimistic, and their cost of capital is actually going down." Right now, the biggest fear about CoreWeave is that their interest expense will grow faster than, you know, revenue itself.
And keep in mind that Nvidia also has an agreement to buy back unsold capacity in CoreWeave until 2032. So they're not doing this lightly, and it's just something we need to remember: Jensen is backing his words with action, not only by investing $2 billion, but also by committing to buying more and more chips from CoreWeave itself.
The third thing here is a super investor who is buying CoreWeave, one might say, sparing no expense, and it's none other than Katie Wood, understand? She took decisive action and purchased $13.5 million in CoreWeave shares, relying on Jensen's words. So now you're starting to see more support from hedge funds and institutional investors.
They said the power— obviously, Jensen said the power is growing every day , and she has significantly reduced her stakes in Amazon and Palantir to fund this. So these are two very, very good companies, but what does that tell you about her conviction? And if we look at some of Katie's holdings, she's obviously a very big investor in Tesla, a big investor in SpaceX, Shopify, Coinbase, AMD, Circle, you know, in their innovation ETF.
They buy many of these companies for the long term because they believe in this development . But she already owns almost 2.38% of her portfolio in CoreWeave, having just added another 13 million, and will likely average the price over the next few days. As we can see, she purchased , uh, 188,665 shares on Wednesday.
She bought another 192,566 shares on Thursday . And today the shares fell another 12–13%. I assume that Katie will be buying more and more shares here . As always, I don't usually advise just following super investors, but we should still keep an eye on what they buy and why they buy it. So, with that in mind, can they actually pay off this debt and won't they go bankrupt?
So they have $104.2 billion in contracted revenue. Their income increased by 112% year-on-year . Revenue grew from 1.2 billion to 2.5 billion in a year, you know? Now, the business also generated 1.5 billion in adjusted EBITDA, you know? So this is not a company that, you know, borrows billions of dollars and tells us that it hopes that demand will eventually come.
They are building the infrastructure to fulfill the contracts that these customers have already signed. So if we look at the percentage coverage here, which will be our first real graph. You can see here that they have an adjusted EBITDA of almost $1.5 billion, which can cover their interest over 2.4 times, you know?
Their net interest expense last quarter was $640 million. I understand this doesn't take into account the amortization of principal payments , but right now we're seeing GPU prices actually go up. This only plays more into CoreWeave's hands. So even if I showed you this graph, it would only reinforce my point even more.
But it shows that businesses here are now generating significant operating profits relative to their interest expenses. So if we want to look further, you know, where this money is going, we certainly see that most of it, you know, goes to real estate and equipment, you know? As you can see, it went from 30.6 billion to, you know, 46.7 billion in just 6 months.
So, what is it ? What kind of real estate and equipment are these funds going to? These are GPU clusters , which are, you know, data centers and infrastructure that are required to convert this type of contracted portfolio into real revenue, you know? So the debt doesn't just disappear into operating losses and get burned.
It actually helps build a huge asset base that we're seeing, you know, being valued higher and higher because of the rising prices of GPUs right now. I wanted to look here, at the third chart, which is our funding chart. As you can see, in the first half of the year, well, CoreWeave raised significant debt capital and, you know, a bunch of equity financing.
We understand that. But this, together with operating cash flow, is sufficient to finance such infrastructure purchases. They ended June with 11.5 billion in net new debt, you know? They raised $3 billion in equity and had, you know, $3.7 billion in operating cash flow, okay? They also had $ 5.5 billion in cash and $15.5 billion in stated liquidity, including, you know, unused credit lines, okay ?
So management basically believes that it has more than enough liquidity to meet any obligations over the next 12 months. And as you can see here, you know, liquidity is 5.5 billion , they have 10 billion in unused credit line. So , they definitely have a huge margin of safety for next year. And here's the last graph I want to break down here, okay?
This is our analysis of what's happening with the interest expense table , okay? As we can see, interest expense increased from 267 million a year ago to 640 million last quarter, and management is forecasting approximately 900 million in the third quarter. But I think people are missing the fact that management reported a reduction in the average cost of borrowing by almost 300 basis points, which effectively translates to $1.1 billion in annual savings.
And if we go back to what Jensen says, he says, "I'm optimistic that their cost of capital is actually going down ." This is a person who constantly communicates with Coreweave because they are one of their partners, he makes purchases from them and understands what is happening. Coreweave announced this last quarter.
And, you know, Coreweave is going to report soon. Well, what do you think will happen? If Jensen says that, I think they're going to show, you know, that their cost of capital actually continues to decline even more, especially with GPU prices rising. But my main conclusion is that Coreweave is not borrowing money just to survive, okay ?
I also want to show you some other neo-clouds and what's going on there. Nebius is a very, very strong company. We see it heading towards the fourth wave at around 0.618 Fibonacci, very attractive, $159 per share. We see companies like Cypher just take a devastating hit, falling from $30 to $9 . We see companies like Hot fall from $140 to $68.
So this is not Coreweave's problem, but we know that the financials and fundamentals support Coreweave's growth, and we see this as an attractive buying opportunity, even though the headlines in the media say otherwise. These are companies that see this demand, this transparency. They talk to the CEOs of hyperscalers every day, and they essentially say, "Hey guys, we understand the situation better than anyone, we're on the ground and we see just crazy demand." And they say he is insatiable.
The demand for computing power is insatiable. They can get funding. They have many such contracts. So I don't expect that to change. Finally, I just wanted to touch on what this bottoming process that we've been observing looks like? First, you want to see high trading volumes and certain levels of surrender.
So when you see Coreweave trading at a volume of 63.5 million shares. If you see the VIX skyrocketing, and we see the volatility index, it jumped almost 30% today, which shows real fear in the market. And if the VIX can reach 25 or more, that's usually a good sign that the market is nearing its bottom, okay?
We have already seen many different scares when the VIX was around 25, like in November. We saw this in October. April was a bit different due to macroeconomic factors. We saw this in March. It is these rapid spikes, when the VIX jumps sharply, that mean that fear is in the market, and the market trades in inverse proportion .
This is a complete washout. You can even see social sentiment in X, on social media. When everything around you screams about a bear market and you feel uncomfortable, it's usually a good sign to buy. And yes, join our group on Patreon. We discuss this every day. We are looking at these companies. We analyze fundamental indicators and give you peace of mind that buying on a dip is normal when there is a lot of fear and anxiety in the market.
That's all for today in the video. I hope you liked it. And if you liked it, please leave a comment below. Don't forget to subscribe , and I hope to see you in our Patreon group. So, take care of yourself, friends.
Comments 0