The Trading Floor: Nvidia Earnings Explained

Hello, welcome back to the show. We have a bit of a breaking news episode and no, this isn't about Oasis Renewables. This is about NVIDIA's earnings. So the second biggest piece of news we've had this week, and that's because NVIDIA's earnings came out last night. The headline read NVIDIA reports 122% revenue growth. But their sales outlook disappoints. Their shares dropped about 8% after market last night, which is a big move for a company of that, that size of magnitude. What we are gonna do in this episode is break down some of those numbers, points of interest. We'll explain not just from the earnings numbers, but also from the conference call. So we'll touch about data centers revenue and growth expectations. Outlooks, and also their upcoming Blackwell chip and how that's going. Then we'll put this into the context of the general macro backdrop and look at equities from a sector perspective. So starting off with the main headline, their Q2 adjusted EPS did beat expectations, $0.68 against $0.64. Revenues also beat. But one thing that you were talking to me before we jumped on here is was data center revenues. So perhaps you can explain what are data center revenues and then why has that specifically as a metric stood out to you?

Well, I mean, these NVIDIA earnings, as we were saying on the podcast last week, and you know, you can read all over the press, you know, are considered to be one of the absolute premier um, economic events of the quarter, right? It's right up there with U.S. labor market data, U.S. inflation data, the Fed, all that kind of stuff, right? And the reason is because of this data center stuff. And that's because NVIDIA is the— absolutely the kind of forebearer of this, or the poster child of the AI revolution. And the reason that's the case is because the stuff they build and sell is what the big tech boys— so Google, Meta, Amazon, Microsoft— is what they need. Their chips, their H100s, their Hopper chips, are what they need to build out their, their kind of data center compute power, which is what they are using to train their, their LLMs, right? And they're, they're kind of AI models. So it's, it's, it's very much this data center revenue from NVIDIA is absolutely a proxy measure for, uh, CapEx expenditure in the AI space. Therefore, it's how much are companies investing in going after this AI revolution. And so it's, even though it's one number and one company and how much they're making in one of their revenue streams. When you follow that trail, it just becomes this absolute bellwether proxy for the AI revolution, full stop. And that's why it's so important.

And so what was it that you interpreted from the data center number from last time?

So there's a few— I think there's a few disappointing numbers in this report. Disappointing. I need to qualify that. This stock, I mean, it's up 160% or was up 160% year to date, right? In just shy of 9 months, it's up 160%. I don't know, go back over 2 years, it's up 1,000%. Even though it's up by that much, you're still, you know, it's enterprise to revenue multiples 24 times. This is a beast of a stock price. Its valuation, well, it got to $3.1 trillion before these earnings, right? And this is all based off sensational stellar growth rates that we've seen in the rearview mirror over the last 2 years continuing into the future. Right. The price of this stock is pricing in phenomenal growth still yet to come. And 120% growth on the top line revenue is— it's great, right? But it's not as good as growth rates in previous earnings reports. So just to put that number, it's 120% growth rate on top line revenue, $30 billion. The growth rate in last quarter's report was 262% growth. So the growth rate has more than halved. That's on the headline revenue. If you, if you go down one level to that really key revenue stream of data center revenue, yeah, growth was, was fantastic, up 154%, right? But the growth rate in the previous quarter was 426%. So the growth rate there is, I mean, well, it's it's not far off quartered. So I guess, you know, as, as the months tick by, the, the comps from the quarter before and the year before get ever more challenging to beat. But when you've got a stock price and evaluation like they have, they've got to carry on beating it. And so the decelerating growth rates for me is the kind of, uh, the really key part of this report, and I think confirms in my mind that NVIDIA is no longer the leader of these broader indices to the upside. The question is, does it become the leader on the downside?

Um, decreasing or decelerating How is that surprising?

Yeah, it's not, but okay, so let's put a number on it. Forward guidance, always a really important part of these earnings, of course, right? Yes, we want to know what's happened in the previous quarter, but, but from a share price point of view, it's way more important, right? What's ahead? What is your revenue going to be in the quarter ahead of us, Nvidia? And so they announced this and their forecast was that it would be 32.5%. Oh, sorry, $32.5 billion. All right. With a plus or minus 2% kind of variance there. Right now, the market had— well, average expectations were $31.9 billion. So their forecast is just a touch above expectations. We're not— you know, Nvidia hasn't been in the business of marginally beating expectations. They're in the business of absolutely destroying and smashing. Expectations. Um, the, but the most bullish on the street in terms of the forecast was $38 billion revenue for next quarter. So they're $32.5 billion. That's well short, you know, that's 20% off. We're not talking small amounts here. And so that was taken, that certainly contributed to the, um, the negative reaction in the share price that we've seen overnight.

And then the other part that, that stood out, so just to get the full picture, was talking about this succession of the technology, i.e., moving into Blackwell. And I know that they had said previously, kind of primed the markets that there'd been some problems, but in this conference call, they actually acknowledged that there was problems. And I like the spin. That they put on it. So just to give it a bit of an idea, they said they had to change a mask production step. Importantly, that's going to improve its yield. And what is a mask? The mask is the template used to burn the circuit pattern into materials deposited on the disk of silicon. So what this is saying is, it sounds to me like they've had a problem. And the way to make a positive out of that is that this is going to radically increase the potential revenue output and production of this thing. But yeah, what did you think of that when you, when you heard about the Blackwell situation?

Well, well, actually, just before I touch on that related, just the final set of numbers, their gross margins, which is kind of related to, to what you're talking about here. It's like that. How efficient is the manufacturing process. And that disappoint— that actually missed expectations. Their gross margins for the quarter just gone was 75.1%, the expected was 75.5%. So that was a straight-up miss. So most of the numbers slight beat, and then that one miss, and that's just not good enough for their share price. So I, you know, on that gross margin thing, well So look, Blackwell is their next-gen chip, and it's supposed to be multiple times more powerful, right? And back in— the thing is, back earlier in the year, they said it would be ramping up to full production in the second half of the year, right? We're 2 full months into the second half of the year now, and still it's not being shipped. Now, they've said they expect to make several billion dollars in revenue from Blackwell chips in the coming quarter. But that's very, very vague guidance right now. We've talked about in the past there's this sort of, um, I can't remember the terminology they use now, but basically there's a gap. If you've got the next-gen chip in the pipeline, it's coming, and you, you know, you're sensationalizing how amazing it is compared to the Hopper chip, the old one, then your Googles and your Metas and your Amazons and your Microsofts, are they going to go, actually, you know what, let's stop buying Hopper, what's the point, we'll just wait for Blackwell? So is now maybe some of this slightly underwhelming revenue beat, maybe it's the beginning of these big boys starting to go, you know what, we're going to stop buying Hopper now, let's just wait for Blackwell. Now, if Blackwell gets delayed further, then that gap widens and it can become a much larger, you know, revenue risk for Nvidia on these forecasts that they're putting out. So obviously there's a lot of unknowns there. Do, do these big tech giants, do they have to carry on buying Hopper chips? You know, do— because they don't want to— there's a, there's a race on here. So it's the biggest race in history in many respects in terms of the money that's being thrown at it. And do Meta want to sit out of the race for 2 months whilst they wait for Blackwell? Or do they go, oh God, we're just going to have to carry on buying Hopper right up to the day that Blackwell actually gets shipped. So we don't know if there's going to be this, this gap or not, but it's certainly an unknown and a risk and just one of those elements that, that adds to a bit of nervousness around Nvidia share price up at these levels.

The final piece within this was the additional $50 billion of share buyback. So is that a little cushion, little buffer to the market to soften the blow? Some of these things you've described?

Yeah, possibly. I don't know what to make of that. I mean, they're, I don't know, they're, they're kind of, I think their cash, um, their cash balance right now is $25 billion, right? Um, that they're not making the profits the likes of those other big Mag Seven. I mean, apart from Tesla or maybe Amazon, who've got super tight margins, but they're not making profits like Google and Meta and Microsoft because they're, they're not as big yet from a revenue dollars and a profit dollars point of view. And yet they're getting the bat out with $50 billion. This is the second, um, share buyback announcement and in, you know, in consecutive quarters. So I mean, yes, it can help, but when you think their market cap is $3 trillion, um, you know, $50 billion, I mean, it's not a lot of money. It's a very weird phrase I've just said— $50 billion is not a lot of money— but it's not. I mean, it's like 2% of the market cap. And so yes, it can help, but I think all the other stuff we've spoken about is way more important for influencing the share price than that stock buyback line item.

You said, you said looking forward is, is ultimately really critical in terms of an investment perspective, direction of where the stock is going. And I, I totally agree with that. And I actually think there's two phases to this. The first phase being Everything you've described— data center underwhelming, bad outlook, problems with Blackwell, margin squeeze— these are all far from the, like, rockstar performance that we've seen for the, for the last year and a half or so, however long it's been. However, I think that this is a really healthy transition for the next leg for for NVIDIA in the sense of it was always inevitable they're going to have to normalize. You mentioned about comps, it's gonna get, that's bound to happen. And so I think they're right in a sense. I mean, the numbers are the numbers, but the guidance where it's, okay, do you know what? I'm not gonna stick a figure on Blackwell. I'm gonna be vague. I'm going to say several billion. The analysts are going to probe me and push me. I'm not going to give you specifics because I don't want to set the bar out. All I want to do is manage. I want to manage the deceleration of the metrics you've described. And actually, I think that they, you know, you say context, the stock was up, what, 300% last year, up another 150% plus this year. They've come off 8%, and the market bounced all of the moves in the NASDAQ futures from last night. So I think they played a blinder here.

Yeah, okay, so I guess that their challenge is, yes, managing the deceleration. To put another way, it's kind of managing an orderly deflation of the bubble, right?

Exactly.

And And yeah, I mean, I think one, the, you know, the first kind of part of that chapter is, right, let's, let's make sure on the guidance side, as you're saying, let's, let's be conservative. We've got, you know, straight out of the Steve Jobs playbook, um, let's be conservative with earnings and then let's go and smash them. Um, but yeah, I, I don't know whether— the unknown for me is You know, this is the end. And actually, Nvidia's share price peaked so far this year, and I think in your mind will beat that, but at the moment it's June the 18th and it was $130, okay, $131.85, okay, that's the high, and it remains the high to this day, right? Now what I want to know is, is this share price up here Does that account for Nvidia no longer absolutely dismantling and destroying all expectations? Because I think they're into the next phase, which is now— look, don't get me wrong, these numbers are amazing, you know, growing top line growth at 120% when you're clocking in, you know, $30 billion. It's, it's, it is amazing, but but it's not rockstar. So I think we're into the amazing phase now. The rockstar stuff's, you know, the Oasis rockstar phase has, has ended. So I'm just thinking, well, where's the share price? What's the right share price for Chapter 2 here? And I, I think it needs to— let me put it this way— I think it needs to grow into this share price. I don't think it'll take long, cuz look, their growth rate is still fantastic. But I think there's a period of calm here, and it'll— and we've seen that in the last 2 months, 2 and a half months. It hasn't gone up, it's gone sideways with a lot of volatility. So I think it needs to grow into this share price a bit, and then maybe later in the year it can, it can justify going again. Um, but yeah, I think we're definitely into phase 2 now.

To conclude then, can you put this into a bit of context from a wider perspective, looking at the broader equity market and macro picture?

Yeah, absolutely. And so, because as we've said, Nvidia, really the poster child of the tech— of the AI boom, um, you know, tech has been leading the rally. If you're looking at broad indices like the S&P 500, for example, then tech has been leading the rally all year. Nvidia just on its own, um, has made up more than a quarter of the S&P 500's gains. Has come from just that one stock. Um, and so that kind of sums it up. Uh, Nvidia is 6% of that index, and yet it's delivered over one quarter of its total gains, right? But since Nvidia's June peak, there's been a changing of the guard, and tech is no longer leading the index higher. And in fact, if you look since June the 18th, Tech is— tech in this, as a sector in that index, is actually the only one of two that are down in that 2.5-month period. And tech is the worst performer, it's down like 4%, okay, in the last 2 months, 2.5 months. Communications is next worst, it's down like 1%. All other sectors are up, and it's actually being led— so there's, there's been a sector rotation play ever since Nvidia peaked June 18th, and money's coming out of big tech. And we've spoken about this before, it's come out of the Mag Seven, it's going into some of these unloved sectors that have been underperforming, especially the interest rate sensitive sectors now that rate cuts are coming. So real estate, for example, that's been the best performer. It's actually up nearly 15% in the last 2 and a half months, right? Because rate cuts are coming. So there's been this sector rotation, and these numbers from Nvidia last night, they don't tell me that's going to change. I mean, I think that tech is going to continue to underperform, and it's no longer going to be that key driver on the upside. And it's just a question of whether all the rest can actually drive a big enough rally for the S&P to make new highs without tech joining in. And tech is, it's like 30% of the index, right? So that's my thesis, is the S&P is going to find it hard on the upside if tech are sat out of the game.

Isn't this just like a time, uh, challenge then? Because if the, if tech can underperform and the market doesn't seriously sell off, and the rest pushing it and the market goes into a whole pattern. Isn't it just a matter of time before then tech resumes some interest again? And then if you have that already baseline covered, the downside's covered by a broader subset of sectors.

Yeah, you're missing the opposite argument though. So here's what I think will happen. Either you're right. Okay, and rate cuts start coming, there's no recession, or it's very mild, you know, the US labor market stays solid, then you actually start to see genuine revenue coming into these other big tech firms from their AI investment. That's what really what we need to see next is the, the ridiculous AI investment that's going on has to start to bear fruit. So if that can happen, right, no recession, AI revenues start to really ramp, then yeah, we're away to the races and we're going to go again on the upside and tech will be all over it. However, the opposite is that the labor market weakens, the Fed aren't quite as aggressive cutting as people want, and then you get that double negative where tech aren't in the game anymore and and actually this rally you're seeing in all the other sectors really starts to lose momentum, and then you get the double whammy of everything's bearish, and then the index can come off, you know, absolutely. So yeah, we're in an interesting hold pattern here whilst we first— well, next up is next week, labor market report from the US, um, Friday end of next week will be the next key kind of litmus test on the US economy and therefore where these markets are going next.

That's a big one, it feels like.

Yeah, absolutely. Cool.

Well, look, great to grab your thoughts very early this morning. So good to get that out fresh off the press. Thanks, Biz. Thanks everyone for listening. And yeah, see you next week.

Cash later. Thanks a lot.