Everyone Is Missing What Nvidia Just Told Us

Nvidia may have delivered one of the greatest quarters in the history of this company, and I don't say that lightly. It might even be the highest of any company. But I've owned Nvidia for a long time, and I'm still extremely bullish on where this company is headed. And somehow, even with expectations already sky-high, these numbers still managed to surprise me. But this wasn't a perfect quarter. Buried underneath the record revenue, explosive AI demand, and another massive beaten raise, there's a few things that I think investors need to be paying attention to. So, let's break down what Nvidia just reported, why I am calling it the best quarter ever, and the numbers that actually matter for what's to come next. This was Nvidia's Q2 2027. It's technically calendar year Q2 2026, but that's just how they reported in case you see any of those numbers. It's not for next year, it's for right now. That being said, at the time that I took this screenshot, we were up roughly about 9.4%. Right now, we're up about 9.7%, so it's a little bit higher, 229.95. And this is off the back of a triple beat. Revenue beat Wall Street's expectations, not just Nvidia's expectations, but Wall Street's expectations by 4.4%, a larger beat than what they usually sustain. EPS also went up 6.22% to $2.22. That was a really good beat as well, and they also increased next quarter guidance. On top of that, we saw 106%. That's what $96 billion worth of revenue brings us, is an accelerated growth rate, which even quarter over quarter was extremely high, of 105.9%. Re-accelerated from whenever China revenue came offline from data center revenue, and we took a big hit from that. We have since re-accelerated. This is really broken down now into three different categories. You have hyperscaler revenue in the data center side, AI cloud, industrial and enterprise data center revenue, like really big neo clouds and sovereign AI, all of that sort of stuff. And then on top of that, we have edge computing, which is essentially everything else. You have their OEMs, their automotives, and gaming, all of that stuff all put together. Whenever we take a look at the growth rates, the growth rates for the AI clouds and the sovereign AI and enterprise customers actually over grew what we saw from hyperscalar customers. Now, the reason that you want to see that is that there was for the longest time, you can see five quarters before that, the hyperscale growth was extremely high. And that can point to a lot of customer concentration. There was still large customer concentration this quarter, but now, because of customers like SpaceX, we're starting to see large breakthroughs from other customers that really weren't on Nvidia's map before. Even edge computing also doing very well. So, continuous growth across all three segments of the business. This is not at just pure revenue and them resting on their laurels. They are constantly reinvesting in the business to try to make sure that their products and the entire product suite is continuing to improve. So, what we're seeing from operating expenses is that they're climbing. Nvidia breaks this out into two different categories: selling, general and administrative, so sales and marketing and G&A. And then on top of that, the really important one is research and development. Research and development took a really big bump up, and it's going to see an even larger tick up next quarter, but right now it's about 64% growth year over year. However, that is large, it's slower than what revenue growth is doing. So, they're spending a lot, but what we end up seeing is that SG&A ends up falling as a percentage of overall revenue. Now hitting a new company low of 1.4%. Even though they're spending more. On top of that, research and development, we're spending a ton more, only 7.3% of outstanding revenue. Just a couple quarters ago that was 12.7%. And then even on top of this, while the stock is doing extremely well, their stock-based compensation is still doing well, even though we're seeing a ton of reports of people taking great head count or hiring people for really high expenses. Stock-based compensation as a percentage of revenue ended up falling to another low 2.1%. And yes, you guys are correct. Nvidia just hit $230 while we're recording this. Maybe this has to do with profitability. Profitability on operating profit. So, after we strip out a lot of those expenses that we just saw, the SG&A and the R&D, we just hit another high in terms of operating margin of 66.2% bringing in $63.7 billion of operating profit. On top of that, we end up seeing a rule of 40 between their adjusted EBITDA and how much revenue growth we're actually seeing. Another new high, 173%. We've seen it higher, but this was during the initial data center boom. So, then it started to fall off, fell to 117% and now re-accelerating once again. Putting it in like the top five biggest names, which is the Microns and SK Hynixes and all the companies with extreme bottlenecks. You usually don't see a $5.5 trillion company in that mix. This ends up continuing to flow down to the bottom line. What we end up seeing is net income almost $60 billion, 62% net income margins. Unbelievable. Part of that, however, was a $7.7 billion mark-to-market gain, which really had to do with SpaceX, their SpaceX investment. They've been booking a lot of gains due to that circular financing that we've been able to see. It's not just been allowing us to boost up revenue, but also the gains of those investments have also been able to win out really big. So, they've been making billions and billions of dollars. Now, it's not all good news and I'll be talking about some of the red flags in this story as well, but I wanted to highlight this as well. Total cash and cash equivalents hit about $99 billion, up roughly about 20 billion quarter over quarter. That so far is a good sign. However, they did also announce that they took out a $25 billion senior unsecured notes, essentially debt. This brought up their long-term debt from roughly, and we're looking at total debt, so this is including their short-term debt, which is roughly about a billion dollars, but this brought up their sustained long-term debt of around 12 billion to up to 38 billion dollars this quarter. Whenever you strip out the overall interest expenses of this versus how much they're making on the cash and much of what they can do with treasuries on the other end, the expenses here don't even make a dent in Nvidia's overall earnings, and you can see that in the net income guidance. This is the hallmark. $108 billion was expected for next quarter. $108 billion. Now, there was a little bit of a hit here. Gross margin only 78% versus 75%, but we're splitting hairs here versus what the company is actually doing, which is growing at extremely high 90-plus percent growth rates or 100% as we saw this quarter. Inclusive of this, we saw Outlook does not assume any data center revenue coming in from China. China now has loosened some regulations in terms of bringing in H200s for companies like Tencent and ByteDance. So, there will be tens of thousands, if not hundreds of thousands of those GPUs going out to China in Q3. That is not going to make a significant significant difference in this number, but it is fun to factor in the fact that we're not even assuming any of this, and we're already at 108 billion. That would, as Wall Street puts it, put us way ahead of the current guidance of 83.5% growth. This was what they assumed yesterday was 104 billion is what they expect for full quarter. This is since changed. The reason why it's changed is every single quarter Wall Street has been significantly wrong in their ability to guide out for Nvidia's overall growth rates. And this quarter was exceptional because we ended up seeing a beat on revenue that we haven't seen since Q3 of 2025. Every quarter before has either been in the range of 1% all the way up to 3 and 1/2%. This time, 4.4% beat on revenue. And then on top of that, we also saw a significant 6% beat on EPS. So, this is why Wall Street's wrong. They also said that with cloud industry backlog now greater than $2 trillion, this This is, by the way, just hyperscalers. CapEx by the top five hyperscalers is expected to reach nearly 800 billion in 2026 and 1.3 trillion by 2027. This is hyperscalers. The hallmark of a hyperscaler really falls into AWS, Google Cloud, and Microsoft Azure. And you could also put in Oracle if you want. What it doesn't include is Nebius, Nscale, Coreweave, SpaceX being probably one of the largest ones, and the myriad of list. Like, I could go down the entire list of Iren and Together AI and Mistral and all of the names that are not included in this projection. They said, "We also want to announce that a company that has been pretty back and forth on whether or not Traanium is going to be their biggest expense in terms of their own custom GPU or ASIC that they want to sell, they're going to be deploying an additional 2 million of Nvidia GPUs starting this quarter through the second quarter of fiscal 2029, which includes Vera CPUs." So, the companies that are building out their own custom silicon, trying to replace Nvidia, they've got a long way to go before they replace them. AWS is also going to offer Nvidia's NeMo Triton open models through their own platform, Blackrock and Sagemaker. Well, Amazon will adopt Nvidia's Omniverse, Isaac, and Jetson stack, which is for their robotics and physical AI suite. So, this is a larger partnership than just purchasing products. This is recurring revenue and a deeper relationship with a company that has been trying to make their own CPUs and GPUs away from Nvidia. They can't get away. So, now, going forward, as we continue to look into guidance, I'm not expecting 108 billion and to be honest, I'm guiding here for 110, but I'm not even a believer in that. I believe it's probably going to be coming closer to 112 or 113 billion, depending on the new Chinese deals that we end up seeing. But, even at 110 billion, which I'll be conservative here to talk about, I'm now assuming a 92% growth rate year-over-year on top of the quarter last year that grew by 62%. That is extremely significant. On top of that, the next quarter after that to finish out the year, this is really important whenever we're talking about annualized guidance is seeing the projection for the rest of the year. The last quarter is going to come in at what I'm assuming at about 125 billion dollars worth of revenue or assuming roughly an 83.5% growth rate. Depending on how we come in next quarter, this might also be very conservative. As my projections before for full year was going to be 105 and then 120. So, I've had to raise them up by roughly 5 billion dollars each and even then I think that I'm being quite conservative. Wall Street, as of yesterday, assumed 105 billion for next quarter and 118. This will also be wrong because they've already guided for their own numbers 108 billion and likely to continue to grow at what they said next year was going to be 70%, not what they expected was going to be 45%. Collette said, "The surge of AI demand is driving driving a global infrastructure buildout, supported by an expanding and diverse set of growth opportunities spanning hyperscalers, AI labs, AI natives, enterprises, and sovereign customers. We now expect to grow revenue by approximately 70% in fiscal year 2028, which means calendar year 2027." They also said, "This is a supply-constrained outlook." If we take my numbers based on my 110 next quarter, which I believe is conservative, and 125, which I think is a little less conservative, but way within the range of possibilities, that really points us to a $411 billion year for full year 2027. On top of that, if you assume a 70-plus percent growth rate, which is what they just called for just yesterday, what we're looking at is a $700 billion revenue year at a high 73-plus percent gross margin. That's going to flow down to net income, and there's going to be a lot of asset gains. Their investments in Nebius, in Coreweave, and in SpaceX, all of these names are likely to put up high marked-to-market gains, which is going to keep net income above 55%, which is what they've held it out for the full year of 2026 and 2027. Jensen said, "We have a lot of visibility now, upstream and downstream. It is the case that we have never forecasted or never guided to a year in advance, even though our demand is much greater than 70%. Our supply allows us to confidently deliver 70%. We are going to continue to work with our supply chain to increase on that." So, he's saying, "If all goes well if TSMC can bump up their ability to create new fabrications, Micron can end up working a little bit faster. All of these companies that we are going to push at the greatest end, can we actually end up exceeding that 70-plus percent growth cuz we have customers wanting these chips? So, this outlook of $700 billion for next year." He's saying, "This is what we can do at a supply constrained level. It's essentially our conservative guide." Wall Street, as of yesterday, was guiding for $573 billion, only a 44% growth rate. You can be skeptical or splitting hairs if you're talking about 44% to 48%. Not too much is going to happen in the way of upgrades for companies that end up guiding only this far out. But, whenever you're talking about the difference between 44% and 70% growth rates on a supply-constrained outlook, you are completely misjudging where this market is to be. And also, let's talk about complete visibility. Me and you are taking guesses as to where this total AI market is going to be headed. We can try to find snippets of visibility in channel checks here and there, and that's what Wall Street's trying to do, and maybe they have better tools. I promise you. I promise you Jensen has greater visibility than any of those analysts that are putting 44%. He is having direct conversations and direct talks with the customers that have to put in purchase orders. He knows it's going to be 70% because he already has the order slip. That being said, he said, "We also see continued demand for our server CPUs, Vera, approximately $20 worth of orders. Based on our customer demand and improving supply outlook, our preliminary expectations is that the CPU revenue is going to more than double in fiscal 2028, positioning us to be the world-leading CPU supplier." So, they want to go from 20 billion in 2027 to 40 billion by next year. That's with the 700 billion worth of revenue that they're currently guiding for. That really brings me to valuation and why I'm so happy to be continuously holding this business. Right now, or at least at the time that I took this screenshot, it was 5.4 trillion. We're currently sitting at 5.5 trillion when I look up at my stock here. We have a forward price-to-earnings, which is based on the old numbers that Wall Street is putting up, not the new numbers that we just ended up adjusting for, bringing up those growth rates to 70% on old numbers. We're at 19 and 1/2 times forward PE. As the days go on and more Wall Street analysts end up revising those numbers to be more bullish, as they see more of that build-out happening, this could fall as low as potentially 15 times if the stock does not continue to move up. On top of that, well, we did end up seeing a lighter than expected free cash flow margin this quarter. It is historically their worst quarter for free cash flow. They also did a lot of share repurchases and dividends, which I'll show off here in a minute. But forward-based, which is looking more annualized out and taking out some of those one-time hits, we are looking at a 20 times forward price to free cash flow, which is in no way expensive versus the general market. And to put that into perspective, we can take a look at this based on the Mag 7. If we look at the trailing Mag 7 names, you have companies like Microsoft, which are sort of tying them in PE ratio. And then on top of that, you have Google all the way at the bottom, but that's based on mark-to-market gains of their investments in SpaceX and Anthropic. So, not necessarily a clear sign of their overall real adjusted earnings. So, that's why we look forward. And whenever we look forward, you see Google much higher, 25 times, but still great valuation on Google there. A lot of these companies are actually in great rates still on a forward price to earnings ratio. At the very bottom, what you end up seeing is that Nvidia is at 19.3 times or 19.5 times in the last screenshot as well, and Meta being some of the lowest names. Meta's not growing at 100% growth rates and even faster on net income cuz net income is growing faster than revenue. So, this is actually looking to be based on next year's valuation potentially the cheapest Mag 7 name out of all of them. By the way, Tesla is stripped out because they're much, much higher in their valuation. Not that I have anything against Tesla, it's just in case you don't see Tesla there, it's cuz they would be in the the hundreds, and way, way above this overall chart. Outstanding shares also saw a really big decrease. You want this number to be as low as possible. It means that they ended up decreasing their outstanding shares by larger and larger amounts. And you can see by this graph in terms of the overall decreases in outstanding shares that they've been buying back more shares now than they ever have previously, which is actually harder to do. The more times that you end up keeping this rate lower and you end up reducing outstanding shares, the adjusted share price should go up over time because there's less of the shares outstanding that are representing the overall market cap. But, Nvidia's stock didn't move as much, so they were able to acquire a lot of shares. They did this with 19.7 billion dollars worth of buybacks, the highest they've ever done in history. And on top of that, they also did 6 billion dollars worth of dividends. So, coupling those two things together, they returned nearly 26 billion dollars of capital back into shareholder hands, whether that was from a direct dividend or by lowering the outstanding shares. Now, I do want to talk about red flags cuz there were a few, like I talked about the 25 billion dollars worth of debt that ended up climbing up, but even that doesn't really hold a pin to the potential real concerns that people are having, whether that's customer concentration or overall inventory growth. I know that this is a concern for some. This is not a concern for me. They have had few write-downs on their overall inventory, whether that's H20s or H200s, and we saw a write-down this quarter as well. I think it was about 400 million dollars. Now, those could still be sold to China, but they wrote them down anyway. What we ended up seeing in terms of overall inventory here was a large quarter-over-quarter increase in inventory of 22.4% representing 31.5 billion dollars worth of inventory. They have to do this to get ready for Vera Rubin. They've done this in the seasonalities before, and why you see these large increases previously, is cuz they have to do that for Grace Blackwell. So, it's not necessarily a terrible sign, but whenever you start to line it up with revenue, you can see that these rates are getting quite high. That inventory as a percentage of overall revenue, if they can't sell it, they're holding on to a lot of stuff that obviously they don't want to be holding on to. They want real revenue, they want them sold. But, it's because of that forward guidance that you can see that the reason why they're building up this inventory is cuz they don't want to be supply constrained. They want to sell this. So, I would look at inventory not as necessarily a bad thing, but almost like a leverage. If they are able to sell it, it's better that they have it than if they did not. It means potential larger future revenue and earnings into the future. They can't sell it, this is a major red flag, but it's not it it's more like a yellow flag than a red flag. Same with this as well. New commitments. We've continued to make strategic commitments across our supply, infrastructure, and our partner ecosystems to capitalize on the substantial growth opportunities ahead of us. Jensen said we partnered with our extensive network of suppliers to secure critical components needed to meet demand for the next several years. Our commitment has increased from 119 billion last quarter, not year-over-year, last quarter, to now 279 billion dollars. And you can see here a lot of this primarily related to the procurement of memory. 92 billion dollars for the remainder of the year, 87 billion in fiscal year 2028, and 88 billion in fiscal year 2029. This is a lot of expenses versus what we've seen previously. So, if for example, they can't keep up these sales all the way through 2029, this is one of those additional red flags. If they can, and they've already adjusted their gross margin outlook to talk about securing this amount of supply, which by the way, securing the supply in the future guarantees that they can have access to this and also control pricing. This is one thing that I think people get wrong about securing this much supply into the future. You get to write really advantageous pricing for yourself if you are truly good at understanding what your customer needs are over the next few years. The problem is is this levers them up if they get the timing wrong. If the orders don't come in for fiscal year 2028, if they don't come in for fiscal year 2029, you still have to pay the memory suppliers. And that's where you can almost double your losses. Not only are you writing down inventory, not only are you writing down your overall memory chips that would just be sitting in warehouses that you can't get off your books, this doubles your losses if you get it wrong, but if you get it right, this guarantees high margin, this guarantees chips going off the shelves, and it's one of the things that I believe that retail investors misunderstand the most is it's not just about having the best chips. It's not just about having customers that want them. It's about having them being able to be delivered on time and at a fair valuation. And this is where Nvidia does the best in terms of thinking ahead that way you can actually give capacity, give chips out at great prices, on time, and ready for the next order. This is why I'm super excited to continue to be an Nvidia shareholder. I've not sold a single share, and it still makes up for over 50-plus percent of my overall portfolio, and honestly, it's going to be climbing into egregious rates if we continue to have these 9-plus percent days. Very, very happy with this overall quarter. It was, without a doubt, their greatest quarter that we've ever seen. And honestly, I'm just an overall bullish investor right now because we've seen so many companies come out with what I believe to be some of the best quarters that we've ever seen because we are in this sort of roaring 20s-esque market where companies like even Salesforce is seeing increased amount of overall revenue and overall growth because of artificial intelligence. There's a lot of spend, but with that spend comes a lot of new productivity gains and new earnings for companies. So, onward and upward with the S&P 500, with Nvidia, and I'm very, very happy with this Q2, and going to be a long-term shareholder at least all the way through 2028 and potentially 2029.