NVDA: strange reaction to Q12019 report?

Saul and I have been discussing NVDA off-board. Saul, sold much of his position, I think mainly because he sees better opportunities. I can’t argue with that because there are some pretty great opportunities out there. One concern was that revenue growth had slowed from the 50+% last year to around 33% from the past 2 quarters. I was not concerned about it because the midpoint of guidance showed a rebound to 50% growth for Q1 2019. I also thought that the past 2 quarters were light in large part to a lack of inventory for NVDA to meet all the gaming demand. The pent up demand that Huang was always taking about. Gaming still makes up the lion’s share of NVDA’s business. And did it rebound, showing 68% growth last quarter. Total revenue was not up 50%. It was up 66%. This kind of growth for such a huge company is enormous. Y/Y EPS growth was up 141% and full year EPS growth was up 77%. The guidance is for 39% Y/Y revenue growth next quarter. I’d guess that they will beat that. So let’s not look at just one quarter because there was a bump in Q1 due to inventory issues in the quarter prior. It looks like NVDA is still growing revenue at around 50-55% on average. Q1 + Q2 is expected to bring in $6.3B and the second half tends to be stronger than the first half. Management said that they expect that again in FY2019. So at guidance H1 grows revenue at 51% over the prior H1. At 50% growth in H2 we can expect total FY2019 revenue of $14.6B. Yeah, it should be a great year again.

Now I haven’t listened to the entire call yet, but I suspect that the stock is trading down after hours because so many people are in NVDA because they link it to cryptocurrency. Q1 crypto-related revenue was $289M. And the CFO says that NVDA expects Q2 cryptocurrency revenue to be about 1/3 of Q1 which would be $96M. This is no doubt highly disappointing to people who bought NVDA because they think they will get rich from the crypto mania. My opinion is that they are wrong that NVDA is tied to crypto, only 9% of total (next quarter crypto revenue will be 3% of total if the guidance is right). Management has said many times that this is not the case. Even Cramer makes the argument. Seems obvious. The P/E is about 41 but as Tinker has pointed out, and I agree with him, the P/E after this year is done will be substantially lower assuming the stock price doesn’t move. $10 adjusted EPS is probably a good guess. The P/E that the company deserves will depend on what NVDA can deliver in FY2020. Can they keep growing at the 50% revenue growth rate? Hmmm. Let’s see what growth opportunities are ahead:

  1. Gaming still strong. I think it can remain strong. There will be cool games that will have more advanced graphics and will require faster GPUs for the gamers to realize the benefits. Real-time rendering! Could VR in games require even faster GPUs? Gaming growth could be very strong for several years.

  2. Datacenter. The AI revolution is still in early innings. Data center revenue grew at 70% last quarter. That was down from more than 100% in the prior quarters. It seems to be slowing but, as its base is now bigger, it will contribute more to the overall growth rate. Datacenter growth is still well above the average growth.

  3. Autonomous. Still small. Still in development. Once the cars start hitting the road in a couple of year, I think we will see very fast growth.

  4. The edge. This is, I think, the biggest of all the opportunities. AI is happening in datacenter now. Some in private (on-prem) systems, but AI is not yet in the millions of devices that it will be in. I think this where the world is heading. There will be all kinds of small, mobile, autonomous machines and they will require a brain. I think NVDA will be the vendor of these brains. Autonomous cars are just one use case of many. This is probably the main reason why I own NVDA.

So if growth is ahead and NVDA will earn $10 in EPS in FY2019 then what multiple should it deserve? I would say at least 30, maybe 35 or even 40. Share count is not increasing due to their buybacks (BTW buybacks will probably be over $1B in FY 2019) and they even pay a small dividend. So by next January, I think the share should be worth $300-350. Maybe even $400. That would be a 20% increase at the low side and if the P/E is maintained then we are looking at 60% upside. My guess is that it will probably end up somewhere in the middle in 9 months from now (absent some large tech or market selloff).

Chris

71 Likes

Chris,

One thing that caught my eye from the call is how they are doing in Inferencing, which is said to be substantially larger than the training market. And still in a much earlier stage.

Traction is also increasing in AI inference. Inference GPU shipments to cloud service providers more that doubled from last quarter, and our pipeline is growing into next quarter. We dramatically increased our inference capabilities with the announcement of the TensorRT 4 AI inference accelerator software

As you know, there are 30 million servers around the world. They were put in place during the time when the world didn’t have deep learning. Now with deep learning and now with machine learning approaches, the accuracy of prediction, the accuracy of recommendation has jumped so much, that just about every internet service provider in the world that has a lot of different customers and consumers are jumping onto this new software approach

More than double sequentially from Q4 and with a growing pipeline. With 30 million servers built on CPUs that should be a long runway.

Also Jensen had some things to say about competition.

With respect to competition, it all starts with the core. The core is that the CPU scaling has slowed. The world needs another approach going forward. Surely, because of our focus on it, we find ourselves in a great position. Google announced TPU 3 and it’s still behind our Tensor Core GPU. Our Volta is our first generation of a newly reinvented approach of doing GPUs. It’s called Tensor Core GPUs. We’re far ahead of the competition. But more than that, it’s programmable. It’s not one function; it’s programmable. Not only is it faster, it’s also more flexible. As a result of the flexibility, developers can use it in all kinds of applications, whether it’s medical imaging or whether simulations, or deep learning, or computer graphics.

As a result, our GPUs are available in every cloud and every data center everywhere on the planet, which developers need that accessibility so that they can develop their software. So, I think that on the one hand, it’s too simplistic to compare a TPU to just one of the many features that’s in our Tensor Core GPU, but even if you did, we’re faster, we support more frameworks, we support all neural networks. As a result, if you look at GitHub, there’s some 60,000 different neural network research papers that are posted that run on NVIDIA GPUs. It’s just a handful for the second alternative. That just gives you a sense of the reach and capabilities of our GPUs.

21 Likes

Excellent, excellent, excellent overview Chris!

Nvidia has always been plagued by short term traders. I expect this to be the case until sometime after 2019. Once Nvidia ends up in every device we own (and Nvidia loves to put their logos on stuff!) then maybe we’ll get more rational reactions to earnings.

Peace,
Dana

4 Likes

Goldman Sachs reiterates its Buy rating for Nvidia (NASDAQ:NVDA) and raises the price target from $275 to $310, a 19% upside to yesterday’s close.

More action: Bank of America Merrill Lynch raises its Nvidia target from $300 to $340 and reiterates a Buy rating.

Analyst Toshiya Hari notes that Nvidia expects cryptocurrency-related revenue to decline significantly in Q2 based on customer feedback.

Hari: “With crypto now de-risked, in our view, the new product launch in Gaming ahead of us, we see positive risk-reward and would thus recommend investors to Buy the stock.”

John

4 Likes

Almost back to even on the day.

I bought some $260 May 18 call options earlier this morning at $3.57…already sold half at $4.80 and have a sell to close order in at present for $6.20.

Guessing I may be highly likely to regret selling at those prices if NVDA ends up at $275 or $280 by next Friday, which would not surprise me a bit.

This latest quarter was outstanding…as Tinker has pointed out, they’re likely tracking for about $10 of EPS on the year, which would be about 26 P/E at the present price…but the earnings are growing at well over 100%, with revenue having grown at 66%.

Maybe I need to go bump that $6.20 up to $7.10.

volfan84
long NVDA

1 Like

With respect to competition, it all starts with the core. The core is that the CPU scaling has slowed. The world needs another approach going forward. Surely, because of our focus on it, we find ourselves in a great position. Google announced TPU 3 and it’s still behind our Tensor Core GPU. Our Volta is our first generation of a newly reinvented approach of doing GPUs. It’s called Tensor Core GPUs. We’re far ahead of the competition. But more than that, it’s programmable. It’s not one function; it’s programmable. Not only is it faster, it’s also more flexible. As a result of the flexibility, developers can use it in all kinds of applications, whether it’s medical imaging or whether simulations, or deep learning, or computer graphics.

As a result, our GPUs are available in every cloud and every data center everywhere on the planet, which developers need that accessibility so that they can develop their software. So, I think that on the one hand, it’s too simplistic to compare a TPU to just one of the many features that’s in our Tensor Core GPU, but even if you did, we’re faster, we support more frameworks, we support all neural networks. As a result, if you look at GitHub, there’s some 60,000 different neural network research papers that are posted that run on NVIDIA GPUs. It’s just a handful for the second alternative. That just gives you a sense of the reach and capabilities of our GPUs.

Not to say that Andrew Left/Citron Research truly believes everything he Tweets, but this section would be something that would be good for him to digest, if he did.

Here’s a link to a post from a few days ago, when Citron tweeted that Google’s TPU 3.0 was a threat to NVDA.
http://discussion.fool.com/citron-questioning-nvda-vs-google-tpu…

2 Likes

Hey Chris,

After my first deep-dive into Nvidia (which tried to DCF Nvidia), I came to the conclusion (along with everyone else on the planet) that Nvidia was… richly valued.

I think Saul is right (who would think?!?). If you’re aiming for fast growth, a large chunk of that potential you see is priced in already. It’s certainly possible that Nvidia will be worth $300+, but it becomes difficult to imagine. Arista for example was much easier to get higher multiples. Even Google appears better value than Nvidia.

My challenge is answering “What do I think that the market doesn’t?”. As far as I can tell, the market thinks Nvidia is in the midst of a bunch of big trends, executing splendidly, and a great company. And all that I believe is priced in already.

It’s certainly possible, and autonomous vehicles are the big possibility. But Waymo is ahead in that game, and they don’t use Nvidia.

Long Nvidia, but not adding.

cheers
Greg

7 Likes

After my first deep-dive into Nvidia (which tried to DCF Nvidia), I came to the conclusion (along with everyone else on the planet) that Nvidia was… richly valued.

I think Saul is right (who would think?!?). If you’re aiming for fast growth, a large chunk of that potential you see is priced in already.

Not everyone on the planet. I don’t think so and those analysts that have price targets above $300 apparently don’t think so.

What is priced in? If earnings grow at 40% then maybe that level is priced in but the guidance is for revenue to grow at 40%. I think 50% revenue growth is more probable. My opinion. Even 40% revenue growth translates to much higher EPS growth which means that a 40 P/E is reasonable (at this high of a growth rate).

But if everyone thought the way I did then the share price would already be much higher. I’ve placed my bet and only time will tell how it plays out…

Chris

11 Likes

Yes definitely you place your bets. If everyone thought the way I do, the world would be full of very annoying people!
:slight_smile:

I’m growing revenue (35% cagr) to $50b in 2022 and getting a price of about what it is now. Obviously, DCF has a lot of unknowns that are somewhat randomly selected, so I only use it to gain some kind of idea about what the market thinks the growth rate will be.

If you think 50% revenue growth over the next 5 years is going to happen (assuming margins stay similar), then the current price would probably be a buy point.

The big unknown is the power of those big trends. If they’re bigger than I (and the market is) giving them credit for, then Nvidia would be worth a bunch more. Smaller? A bunch less.

Can Nvidia grow revenue to $50b in 2022? Around that is what the market thinks from my (not to be trusted) calculations. 50% CAGR would be around $83b.

I think thats why the price didn’t move much. As far as I can tell, it was an excellent quarter, way above average Q1 which is normally a weak quarter. But I believe the market is expecting excellent quarters from Nvidia and thats baked in to the current price.

cheers
Greg

5 Likes

And all that I believe is priced in already.

And then there’s the possibility that the crypto-currency exposure is really much bigger (i.e. because of temporarily higher GPU margins) than the USD 300 million from dedicated mining cards.

2 Likes

I’m growing revenue (35% cagr) to $50b in 2022 and getting a price of about what it is now. Obviously, DCF has a lot of unknowns that are somewhat randomly selected, so I only use it to gain some kind of idea about what the market thinks the growth rate will be.

If you think 50% revenue growth over the next 5 years is going to happen (assuming margins stay similar), then the current price would probably be a buy point.

Greg,

Certainly a valid assessment and way of analyzing. However, even 35% revenue growth translates to much higher EPS growth which would make the current 41 P/E not overvalued in my opinion. For FY2019 there’s no way revenue growth will only be 35% growth. Not with 66% growth for Q1 already locked in and guidance for Q2 at 39%…that’s already 50% for H1. The question really is what will be growth in FY2020 and beyond. You say 35% revenue growth. Maybe. I think higher but we will see…