Nvidia is Expensive, But Take Another Look
I'm starting to see a lot of Nvidia analysis that goes like this:
  • Data center revenue growth is slowing down
  • Competition from AMD and Intel is heating up
  • 17x forward sales is simply too big of a premium for the stock

For me, Nvidia is starting to look like one of my favorite types of investments:
A great business that is currently priced high because of one revenue driver, but hasn't priced in a different revenue driver that will take longer (much longer- think 10 years) to materialize.

The Focus Is On The Wrong Revenue Driver, For Now
Nvidia's stock is up 127% year to date. That's fantastic. Wall Street didn't expect the gaming segment to be up 37% year over year. And they don't expect fantastic growth in gaming going forward. On that point I am in agreement.

So where will the growth come from? Data centers? Also no. Nvidia projected that a slowdown in sales in the fourth quarter will cause data center revenue to contract sequentially for the first time in more than a year and a half. Wall Street is pricing this in too, which is why the stock has stayed largely flat the last four months.

So what is the future revenue driver that makes me want to own this stock for the long haul?
Post media
Artificial Intelligence
The main thing to know about artificial intelligence is that it is expected to add $13 trillion dollars to global economic activity by 2030 (Source: Mckinsey).

Christopher @chasinggreatness wrote a great primer on artificial intelligence if you want to dig deeper.

It is still very early for artificial intelligence adoption. The new technological capabilities enabled by AI are:
  • Computer vision
  • Natural Language Processing
  • Virtual Assistants
  • Robotic Process Automation
  • @tyler let me know if I'm forgetting any πŸ˜‚

Businesses thinking about adding these capabilities today are faced with stiff investment costs which will negatively affect cash flow in the near term. But companies that take the plunge should see an increase in positive cash flow of up to 120% (also according to Mckinsey).

The point? The huge $13 trillion AI opportunity is not priced into the market yet. And Nvidia is positioned perfectly to take advantage of it.

My thesis is that Nvidia is priced for perfection for its gaming and data center segments now. But in the next 2-3 years it will start showing off what it can do in revenue from AI use cases. And that will dwarf its current main revenue drivers.

I am a long term investor by nature. I like to buy and continually verify. Nvidia is a stock I have owned since 2017, but its also one I'm looking to open up a much bigger position in to hold for the next decade.

The way I see it, Nvidia has a shot at being one of the most valuable companies in the world by 2030. With that time horizon, and the likely S-curve adoption that AI is likely to go through, it's much better to be early and just hold with conviction than it is to try and time the perfect moment that Wall Street starts to realize the potential of AI.

What are your thoughts on $NVDA ? Do you like AMD more? What are your reasons?
Amanda's avatar
I sold $NVDA in March to buy $ZM $PYPL and a few others... It’s been on my mind that I might want to rebuy it. Great memo!
Nathan Worden's avatar
It's one of those companies that will almost always have a legitimate "it's too expensive" argument to go with it. But it seems like it can be a juggernaut in AI. And that isn't priced in yet.
Aquiba Benarroch, CFA's avatar
I agree, NVDA is one of those names that I like but always think it's too expensive. And from what I've read, the acquisition of ARM will make it an even more formidable rival, one where $INTC and $AMD will struggle to compete with even more.

I think I should take a closer look as well. Thanks for the writeup @nathanworden
Nathan Worden's avatar
Yeah the acquisitions have been an interesting part of the story here. ARM, Mellanox, and Cumulus.

And the competition is doing it too: AMD acquiring Xilinx, and Marvell acquiring Inphi. They're all trying to set themselves up to remain competitive in AI.
PAUL  PETRO's avatar
I think NVDA is stock of the future. Once its acquisition of Arm goes threw. I think it will start running again.
Nathan Worden's avatar
I agree- the Arm acquisition is interesting.

I see that you have about a 3% allocation in your portfolio to Nvdia @paulpdc , are you thinking of increasing that allocation in the next 6 months or so?
PAUL  PETRO's avatar
Yes, Not in a hurry to buy right now. Like it at under 500.00. But feel it wont pop till Arm is approved. I bought a little Rocket today under 20.00.
Nathan Worden's avatar
Yep, I'd say having Nvidia on your watch list, ready to jump on it under $500 is a good strategy for now.
nicely articulated ,my avg cost is around $500 and I am planning to hold my core position

Nathan Worden's avatar
Yep, its a good one to hold.
Ambrose's avatar
Honestly, I think people overestimate or underestimate NVDA. I think its certain that NVDA will definitely be one of the largest contributors to the AI industry in the future.

However, I don't think its going to be taking a large portion of the AI industry revenue.

I think a large portion of the AI industry revenue will go to companies creating the software product and selling to customers B2C than selling the hardware itself. Additionally, other companies are also creating their own equipment too like Apple CPU chips or Tesla GPU for autonomous driving.

Nevertheless, I still foresee NVDA will continue selling high-end GPU and equipments in the future, good revenue growth, merging with companies to maintain growth, and trying to increase their range of products (AI equipments/GPU). Kinda like the other megacap tech stocks.

So, my evaluation for NVDA is that if you have it good. If you don't have it, nothing to panic about. NVDA is not a 100x stock. But its a multi-bagger in 10 years like the other tech stock.
Nathan Worden's avatar
Ooo that is an interesting take I haven't heard yet.
Tyler Lastovich's avatar
Another small feather in its cap is that Nvidia is a major player in crypto mining. They sold $175M worth of the brand new 3000 series cards directly to miners in Q3. The latest crypto bubble should help drive that even more - at least in the short term. Sadly this is less relevant for Bitcoin (which is mainly mined on ASICs), but ETH should still be fine on NVDA chips for a few years yet while v2.0 gets going.