저자: Kevin · 외국인
요약 보기
OpenAI 주요 투자자(아마존·엔비디아·마이크로소프트)의 투자 축소 움직임을 근거로 OpenAI 기업가치 830억 달러의 신뢰성을 의심. 아마존 500억 달러 투자는 IPO 또는 AGI에 연동되어 150억 달러로 줄어들 수 있고, MS는 '100억 달러 미만'에서 '저(低)수십억 달러'로 표현이 바뀜. 엔비디아는 현금 약 1,000억 달러에도 공급 약정 952억 달러와 단기부채로 여유가 부족하다고 분석. 아마존은 현금흐름 악화로 차입 중이며 MS도 잉여현금이 약 108억 달러에 그친다고 주장. 투자 축소가 주가 하락과 시장 공포로 이어질 수 있다고 봄.
다듬은 글
smokes are the rich backers of OpenAI, including Microsoft, Amazon, and Nvidia, potentially scaling back their investments into OpenAI because they're starting to run out of money. This is really bizarre, and I know this sets a high standard, and look, if I leave you hanging on this one, you deserve to unsubscribe.
But, I actually think what I'm going to show you is going to make you scratch your head and go, "Oh, damn. They are starting to run out of money. And I thought I thought Nvidia had all the money in the world." That's what I thought, too, until I saw what I just saw. Let me show you. First, let's start with Amazon.
Amazon, on January 29th, the day after my birthday, wrote that they were in talks to invest up to $50 billion with OpenAI. This was an exclusive, as reported by The Wall Street Journal, and you can see here, "Amazon in talks to invest up to $50 billion." This is as part of ChatGPT's and OpenAI's new round that could value it as much as $830 billion.
Notice there are no strings attached here, and keep in mind that valuation of $830 billion. Well, something really interesting just happened, and it could signal a collapse at OpenAI, and that its rich backers are starting to run away. Watch this. Amazon's $50 billion investment into OpenAI now could hinge on IPO or AGI.
In other words, attached. Hey, either go public so we could validate that your stock could actually hold up on its own, it's actually worth what you think it's worth, or figure out how to get how should we put it artificial general intelligence, which we're not close to. And until you do that, we're actually only going to invest 15 billion dollars, maybe, because that's just still being negotiated.
This is starting to sound a whole lot like the walk back that Nvidia just pulled on OpenAI. Remember when Nvidia had that joint press release on the OpenAI website where they both agreed, "Hey, we're going to deploy 1 gigawatt of power. Every gigawatt that's deployed, we're doing a 10 billion dollar investment, and we're going to do it 10 times or up to 10 times, that's the goal.
The total value of this is 100 billion dollars." And everybody was so excited about how Nvidia and OpenAI had this massive circular partnership, mind you. And then, of course, Jensen kind of rug pulled that and "Oh, no, no, we never had that idea. Uh you know, we were invited to invest to that much, but no, I mean, we never made any formal agreement to that." And now, they still can't even conclude this 30 billion dollar round that's being talked about, which maybe isn't surprising because once we start looking at the Nvidia balance sheet, we might understand why.
But first, understand what's noted here. The valuation for OpenAI is now sitting at 830 billion dollars, quote, "before the financing." Okay, that's interesting because now they're saying Amazon might invest up to 15 billion dollars in this round that'll value OpenAI at 730 billion dollars before financing, okay?
But then, here says that the round could value it as much as 830 billion dollars. Now, it's possible that this is where they're talking about after that 100 billion dollar injection because the numbers obviously align perfectly, but I think what they're trying to now say is they're conveniently leaving off over here that this valuation includes that $100 billion.
And if that 100 happens to only be 70, then I guess the combined number is actually going to be 800. Or if they raise less, then it's 780 or 750. And all of a sudden, that OpenAI valuation that felt like it kept going isn't actually really growing that much. It's only growing based on the very new dollars that are being added.
And maybe that's why companies like Amazon and Microsoft are like, you know, maybe we don't really want to throw any more money at this piggy. In fact, Microsoft is now no longer thinking about investing Well, they were they were thinking about investing quote less than $10 billion. So previously, it was being framed as less than $10 billion.
That is now being phrased as Microsoft is expected to invest in the low billions of dollars. So we can see like the framing that you get. You know, you hear less than 10, you think like eight, right? Now we understand it's all framing and marketing. So it technically could still be accurate. But framing and marketing does a lot when people write these things or leak this information.
Now Microsoft is expected to invest low billions. Well, that could be like 2.4, right? That is that's way smaller of an amount. And if Microsoft is pulling back, and Nvidia's pulling back, and Amazon's pulling back, the whole round is starting to actually look like it's fizzling. Which isn't good. Is it potentially because these companies are running out of money?
Well, let's just understand really quickly what's going on in some of these balance sheets. I think the first one, because we really propped it up this much, is Nvidia. Let's go there. Ah, excuse me. Still have this lingering cough. It's like the most annoying thing ever. But anyway, so first, we'll go over to the Nvidia statement.
And look, Nvidia has a glorious amount of net income. We know that. $42.9 billion of net income. So, they should have plenty of money, right? I mean, after all, cash from operating activities, $36 billion just in the quarter. Now, in fairness, they did acquire Groq for $13 billion. That helped increase their goodwill substantially.
So, they did use money to make, you know, equity investments into not just Intel, but other companies that sort of help like startup investments that they make. They disclose that in their 10-K here as well. Uh and then, of course, they invest into treasuries and to have cash available for their supply chain commitments.
And so, more importantly, I think it's looking at the balance sheet. So, if you look at the balance sheet, you can see that Nvidia has cash and marketables plus accounts receivable. I'm going to put these all together of a hundred billion dollars. They've got 21 bill of inventories. Okay, great. So, we've got a, you know, a hundred billion dollars of cash and accounts receivable.
Seems like they got plenty of money to go invest $30 into Open AI, right? Not necessarily. We've bills to pay of about $32 billion over the next year. So, that $32 billion is a problem already because it means now we really only have about $70 of free cash. But, let's just assume for a moment that our current liabilities are just going to wash out our inventories, okay?
So, then we got a hundred billion dollars of actual cash sitting around. Well, we could go throw that at Open AI, right? No, not so fast. There's a little bit of a problem in a little footnote that we get in the CFO commentary, and it happens to be right here. Balance sheet and cash flow. The total supply-related commitments that Nvidia has are 95.2 billion dollars.
In other words, in order for them to actually manufacture the product that they need for the next quote several quarters, which is probably three quarters, right? Just to get through the the rest of this calendar year probably or their fiscal year, roughly the same. They need to spend 95.2 billion dollars.
So, they need to start sending that money as deposits. They still need to collect their accounts receivable. They need to pay their short-term debts. They want to continue to make investments into whether it's CoreWeave or other startup companies. And at the same time, they got to not only pay for the deposits, but keep paying the bills when the chips actually get manufactured and the deposits to secure those supply chain commitments.
That's how they get the lock on this industry. Trust me, not everybody gets to make 75% margins by designing a chip. Nvidia doesn't make the chip. Nvidia doesn't really do anything other than design the chip. Now, that's very important. Don't get me wrong. But really what they're doing is they're a middleman.
They reserve the manufacturing capacity and pay a deposit. They collect the orders for the chips. Then they have the chips made, shipped to the customer, and then they collect the money. It's a really brilliant high-margin business that a lot of people want to get their hands on. But Nvidia needs that capital to reserve and squat those supply chains.
So, by squatting on the supply chains, when OpenAI says, "Like I want that 100 bill." Nvidia's like, "Honey, I make a lot of money, but not that much." So, Nvidia's balance sheet doesn't look that ready to go pour a bunch of money into OpenAI. Maybe that's why they're trying to kind of push this commitment down the road.
Because in fairness, they do generate cash flow about three 30 billion dollars a quarter right now. So, could they technically just use that commitment of 100 billion that they have in in, you know, liquid and their inventories and their accounts receivable? Could they just use all of that and this next quarter's cash flow and just call that their commitment and then maybe two quarters from now invest in Open AI?
Yeah, probably. But, it shows you that there are decisions here that Jensen has to make. Do I squat on inventory or do I give it to Open AI when Claude and Gemini are starting to eat their lunch? Now, Amazon is not much of a different story. Take a look at Amazon. This is Amazon's cash flow statement 2025 for the year.
Amazon's cash flow literally tanked. Let's throw an arrow in here so you can see this a little bit more clearly. See right here? 32 billion dollars of cash flow. See right here? About 33 billion dollars of cash flow. Last year? 7 billion dollars of cash flow. In fact, they had so little cash flow, and this is Amazon, mind you, that they had to borrow short-term debt to refinance some other short-term debt.
They took out more, so they took out net roughly about a billion dollars of short-term debt. And then over here they took out long-term debt of net net, let's call that probably about 8 billion dollars in long-term debt. So, Amazon's actually having to take on a little bit of short-term and about 8 billion dollars of long-term debt because their cash flow is so low.
And so now Open AI is here with their hand out going, "Hey, where's my investment?" And these companies are like, "Yeah, you know, we're kind of running a little tight ourselves." And it is a little concerning because I feel like the whole cycle relies on this. Look at Amazon. Like, "But, Kevin, Amazon has cash of $78 billion on their balance sheet and marketables of 22.
Kevin, if I add that together, that's $100 billion. Yeah, that's great math. But what do we have in bills to pay? Oh my goodness, accounts payable 94 plus accrued expenses 67. That's $161 billion of bills. So they have $161 billion of bills on their desk and they have $100 billion of cash. The math ain't mathing.
No wonder Amazon's like, um man, maybe we got to put some strings on this because we can't afford to make this kind of big investment in OpenAI, especially if it's going to be a freaking loser. Now, if it's going to, you know, go public or if they got AGI, they got something special for us, it might be worth going into debt for.
But otherwise, I don't think we got the money for you, Sam Altman. Sam Altman, let's try Microsoft. Like my maybe Microsoft's better. I didn't actually look at Microsoft yet, but let's just do it for giggles. Let's see what Microsoft has. This is the last Microsoft report. So we're going to go to the last 3 months 2025 cash from operations.
I got 35.7 billies right here. Plant, property, and equipment, wow. We're also down to about $6 billion of free cash flow. That's it. Cuz as soon as I subtract additions to property and equipment, we only have $6 billion of free cash flow. What did they end up doing with financing? Financing right here.
They did end up repurchasing stock, paying dividends, and they paid off some debt. So in fairness, they didn't take on debt like Amazon. Amazon had to take on debt. Now, if we go look at their actual cash available, oh, well, that's unfortunate. I've got cash available of about Excuse me, $24 billion.
[clears throat] Short-term investments of 65, that's 89. But wait a minute, how many bills do I have to pay? Well, let's take a look here. If we go over here, if I take 89 minus 37 minus 4.8 minus 10.1 minus 2 minus 24.3, I've got about 10.8 billion dollars spare uh once I pay my current bills due this year.
So, Microsoft only has 10.8 billion dollars spare. No wonder they are only thinking low billions as they only have 10.8 billion spare after paying their bills. Mind you, they're also sitting Oh, look at all this debt right here. Long-term debt, another 35 billion, 27 billion in long-term income taxes.
These companies are starting to get tight not because they want to be mean to scam Altman, but because they have to start making decisions. And that's something we haven't seen over the last, well, frankly, few years. Because these companies were so flush with cash, they were able to keep the cycle going and they could say yes to everything.
Now, the companies are starting to have to make choices. And so, you're starting to see strings get attached. But the problem is, as strings get are getting attached, companies are starting to go, "Oh, there's less money going around. Maybe we need to be a little careful here. People will rein in and stock prices fall.
People panic." And then this is where Jensen goes on CNBC and he makes this argument that, "Oh, don't worry. Uh software companies shouldn't sell off. Software companies shouldn't sell off because AI agents, well, they could just go use the software instead of humans. And like, why go make a new software?
Just use that software." Man, I haven't heard an argument like that one before. And so, what's likely going to happen is that agents won't replace the tools, but agents will use tool.
Comments 0