Amazon's structural advantages in cost of capital and infrastructure position it as a long-term winner and a lower-beta alternative to high-growth AI stocks.

All right, number three stock that we want to talk about is another one of these huge players. This is This is a theme that a lot of these companies are some of the biggest companies in the world, but Amazon, not [snorts] only one of Nvidia's biggest customers, but again, another one that's trying to get out from under Jensen Huang's thumb.

Tranium is the chip that they have, so that's going to be kind of analogous to the TPUs at Alphabet. And Tranium is one of those It's a little bit behind. They have not been building it for a decade like Alphabet has.

But, the data on it more recently has been that they have been improving extremely quickly. And I think the thing to think about with Amazon is that the incentives are very different.

So, they don't need to be a leader like Alphabet does in building their own models. They can be more of a commodity supplier. And if this ultimately does become a commodity business, then the question is going to be who has the lowest cost infrastructure?

So, you take Nvidia's gross margins from their chips. That's an expense that you're not going to have if you're going to be using your own internally developed chips. Uh they've also been extremely efficient.

They've been doing this for a very long time, so they they have low very low cost of capital. They have this huge retail business that just, you know, happens to be sort of a flywheel that they can also kind of lean on.

But, that's where this power of AWS and the way that they built the the the product and the infrastructure, I think is going to be a huge advantage for them. And oh, by the way, they've already got a ton of data for, you know, millions of companies all around the world.

So, you add all of that up, and I think that Nvidia would like to be the monopoly player in the GPU business. But, when you have huge company a huge companies like Alphabet and Amazon as your customers, they're trying to, you know, kind of commoditize their suppliers.

And I think this is one of the companies that has done it over and over again. I think we'll see the same thing here.

And with a company like Amazon, too, I think there's just so much optionality and it all really comes back to what is the cost per token, you know, I I think that's probably the best way to kind of find a proxy.

So, if you just look at them versus a company like CoreWeave, for example, one of the biggest Neo Clouds, Amazon's just simply going to have a lower cost of capital. So, CoreWeave may sign a big customer and it you know, they may have that may that project may be profitable, but if their cost of capital is, you know, 12% for debt, equity is going to be, you know, 15 to 20%, you know, because you're going to investors are going to want a premium on top of that.

Amazon's cost of capital is much, much lower, half that, if you will, even if they're even if they're taking out debt. I do want to point out that this is a company that is increasing their debt load as they build out a lot of this AI infrastructure, but I think that cost advantage is something that these newer, smaller companies are just not going to be able to compete with.

You add to that the the that they already have, the momentum that they already have in the business, and T uh Tranium on top of it, I I just think this is going to be one of the companies that I've kind of overlooked even over the past, you know, 6 months to a year, and as the stock has gotten maybe a little bit more attractive, I just think that it's going to be a juggernaut for a very, very long time to come.

All right, I want to end with a little bit on valuation because you look at the bigger companies that we talked about, Alphabet, trading for about 25 times forward earnings estimates, Amazon trading for about 27 times forward estimates, AMD and Marvell in a very different boat.

Uh AMD trading for about 47 times forward estimates and Marvell at about 43.

Well, maybe I should answer it like this. I mean, I think that you should expect higher beta or higher volatility with AMD and Marvell compared to Amazon and Alphabet. So, if you are looking to make a single investment today, maybe you want the lower beta stocks in Amazon and Alphabet and you're not too worried about a huge pullback because these are huge powerful companies.

If you're more concerned about the volatility, maybe Marvell and AMD are ones that you dollar cost average in, understanding that chances are they will be significantly lower at some point and also significantly higher.

You don't know where the top and the bottom is. So, you're just going to dollar cost average a position.