What Could Replace Nvidia? Why Is That Important?

Wild swings in equity prices speak to high valuations and uncertainty about whether even the best performing companies can deliver but that isn't all.

There is also the competition factor.

We worried about the issue of market concentration last summer and into the autumn. You can read about it here and here but the short version is we were nervous that the market would struggle to head up if the rally didn't broaden.

Now, however, the reason for Nvidia's fall doesn't have to do with falling demand for their products. To the contrary, all the signs from the current earnings season are that there continues to be even crazier amounts of demand for Nvidia's precious GPUs.

The reason that Nvidia's stock - not order book - has suddenly hit the skids is that constantly underrated economic principle: competition.

The precise reason is that all the biggest buyers of Nvidia's products: Google, Meta, Amazon etc. have announced plans to develop their own "silicon." That means they plan on designing and then developing their own Nvidia-like processors to help them with inference and then training and then finally operating models similar to what they use Nvidia's famous GPUs for today.

This has dramatically shifted expectations about the future size and profits of Nvidia's business. Because even if the chip designer stays ahead of the competition those customers will nonetheless try and move as much as humanly possible away from Nvidia's products. As the demand falls, not only will the revenues fall but also the company's profit margin.

That is the real key to the ever higher stock price: not just selling MORE but selling MORE at ever higher profits.

Here is a recent interview where Mark Zuckerberg discussing Meta's effort to build their own silicon at some length:

There is a simple driver of this: fear. The giant tech companies are in fierce competition with each other and are terrified of losing out to a rival. There is a reason that one of the most famous books to ever come out of Silicon Valley is "Only The Paranoid Survive."

They are deeply uncomfortable being so dependent on a single technology company. Why? Because being dependent raises the chance of being outmatched by another company if and when they develop their own silicon in house. The tech giants might hate the fact that Nvidia can charge so much thanks to their virtual monopoly even more.

That isn't an exaggeration. Nvidia's gross profit margin has been well over 60% and increasingly even flirting with being over 70%.

That is nuts. Many companies fight and claw to protect 2-4% profit margin. 10-15% is considered very healthy. It is tough out there!

These new attacks from rich and highly motivated competitors means the expectation from investors are shifting. The worry is that Nvidia moat around those profits could be closing, perhaps far quicker than people assumed. Investors may be unsure when exactly these companies will have other options and they are also uncertain about just how quickly these competing products are being developed.

So, anxiety around whom and uncertainty around when.

What can you do about this?

There are broadly two options:

  • The first is to spread your bets back among other chip designers, including those racing to develop their own like Meta (nee Facebook), Alphabet (Google), Amazon etc.

  • The second is to buy companies down the supply chain, mostly in Asia, that help Nvidia (or Meta or Google) actually either help or build the processors in question.

Each group of company has risks! The first collection of firms are obviously very expensive and, as we have highlighted before, are vulnerable to more than just their own performance. The likes of Meta could both fail to adequately develop their own processors as well as suffer from being the largest and most expensive stocks in the world's best performing market.

When it comes to the Nvidia supply chain buying mostly Asian supply chain companies and many of them smaller enterprises is not for the faint of heart even before we get to the very real fact that they are clustered around Taiwan (or South Korea!). These countries have serious strategic question marks that are also likely impacting valuations and expectations.

The most important point though is that the artificial intelligence theme has not run out of steam just yet. As Zuckerberg also discusses in that interview, Meta is planning on spending tens of billions in the coming years to make sure they do not only compete but hopefully dominate this space.

Put differently, we have perviously recommended Nvidia as a shovel marker of the AI theme with the argument that in a gold rush it is better to own selling shovels than being a miner yourself. The American chip designer makes a very sophisticated shovel and for a long time, they were the only game in town. However, that is beginning to shift and that will begin to change a lot, in the stock market and beyond.

As we have covered in our introduction, that isn't the only constraint out there for the AI theme though.


Have questions? Care to find out more? Feel free to Download our App (!!) or reach out at contact@pebble.finance or join our Slack community to meet more like-minded individuals and see what we are talking about today. All are welcome. You can also get our newsletter as an RSS feed.


    How To Think & Invest Around Energy Constraints?


    Is The AI Bubble Bursting? Is It Even A Bubble?