Dear Reader,
Anthropic, the most powerful AI company on Earth, goes public today.
It starts trading under the symbol ANT.
That’s going to be very interesting, but I know we can’t talk about that anymore.
There will be a rebellion among our customers if I don’t discuss Nvidia.
We got almost a thousand questions about Nvidia in the past week.
Everybody wants to know, and I said I’d talk about it.
But I figured it was silly to do that until earnings came out and gave us some context.

Nvidia’s earnings were great.
And what they mean for the AI trade is a lot, actually.
Everybody else has given you their 24-hour, very quick version.
They’re great, they’re blockbuster, and so on.
I like to do a little more analysis and tell you the context.
The Numbers
Third quarter revenue was $115.8 billion.
Wow.
That topped estimates.
Folks were looking for $107 billion.
They forecast $130 billion in the first quarter of 2027.
It’s really unbelievable.
One of the things I always tell you to look for is gross margin, because it tells you about pricing power.
How strong is a company’s competitive moat?
Their gross margin is 71.5%, so very, very, very strong.
They still have pricing power.
They can still sell their products at high prices.
And they generated $43.77 billion in free cash flow.
Why I Look at Free Cash Flow
Free cash flow is cash flow minus capital expenditures.
Don’t look at profit when you’re looking at a business.
Profit can be manipulated by accounting tricks.
Look at cash flow.
More specifically, look at free cash flow.
It’s the amount of cash that gets to go into your pocket if you own the business privately.
That’s a very important distinction for folks to keep in mind.
Capital expenditures are the property, plant, and equipment, the things you have to do to make the product you sell.
Plants. Equipment. Factories.
What’s great about Nvidia is that they don’t really have factories.
Nvidia doesn’t make chips.
They design and market chips, very much like Nike.
Nike doesn’t make sneakers.
They design and market sneakers.
Taiwan Semiconductor manufactures Nvidia’s chips.
Taiwan Semiconductor has very different economics.
It has lower returns on capital, because all of that manufacturing plant sits on its balance sheet.
That increases the investment and lowers the return relative to it.
My First Impressions
So what are my initial impressions?
Geez, this is a $3 trillion, $4 trillion company growing at the pace of a startup.
That’s funny.
It tells me the AI build-out is accelerating.
It tells me they’re very dominant in the global data center chip market.
They have 76% of global server GPU revenue, which is three times higher than the combined revenue of their competitors.
Broadcom has 10%.
AMD has 3%.
Nvidia is the monster in global server GPU revenue.
It’s not even a debate, really.
So Why Is It So Cheap?
The most surprising thing is when you put the business next to the valuation.
The company is trading at 10 times forward earnings.
Ten.
It’s just unbelievable.
Why is Nvidia trading at 10 times earnings?
There, my friends, lies the tale of the tape.
First of all, 10 times earnings is what you trade at when you have a scandal.
A great company with a scandal trades at 10 times earnings.
That’s what happens when great companies have scandals.
They crash.
They trade really low.
Smart people buy them.
I like Nvidia down here.
My goodness, it’s cheap.
I don’t have another opportunity for you trading at 10 times earnings, growing at 70% a year, with this kind of free cash flow.
You’re not going to find one.
Enablers to Adopters
The other reason the stock is trading so cheap is the one I talked about a couple of weeks ago.
The AI trade is moving from enablers to adopters.
The build-out trade, the enabler trade, chips and data centers, is like yesterday’s news.
Folks are saying the build-out will be contingent on the people who can monetize it and get a return on that investment.
It could be great right now, but over time it will depend on companies like mine, or like the S&P 500, that take these tools and get a return on them.
One of the big changes is that token prices have fallen 50% to 60% in the past year.
So it looks like the models are going to shift from selling tokens to selling tasks.
What we want to see, as investors in aggregate, for the AI trade to bounce back is UnitedHealth.
We want to see a lab, a hospital.
We want to see companies say, we are using this, not just investing in this.
We are seeing our margins and our profitability go up because of AI.
That’s what makes the build-out sustainable.
There is concern that it isn’t sustainable.
I think anybody who doesn’t believe the build-out is sustainable doesn’t use AI the way we use it here at Behind the Markets.
We are power users of AI.
We don’t mess around.
Ultimately, I think it filters through the S&P 500, based on how aggressive and smart and on the ball their managers and CEOs are.
I think Anthropic and ChatGPT are changing their model from selling tokens to selling tasks.
As that happens, you’ll start seeing them create agents that do specific tasks for companies.
I wouldn’t want to be in the cybersecurity business, that’s for sure.
These guys are creating the things that hack.
They’re creating the hackers.
Of course, they’ll also be able to create the solution to the hackers.
It’s really, really interesting.
The Bottom Line
The AI trade is moving from enablers to adopters.
Until that transition happens, you’re going to keep getting great opportunities like Nvidia.
To me, Nvidia looks like the best cheap stock on the market right now.
Have a wonderful Thursday.
I’ll see you tomorrow.
“The Buck Stops Here,”

P.S. I have an urgent update about Tuesday's webinar.
Our #1 Private AI deal was so popular that it is now oversubscribed.
That means this Pentagon, NASA, and Space Force partner may not even be able to accept your investment anymore!
You may still be able to get in before the window closes, but that window is closing even faster than we expected.
This may be your last chance ever.
Found this helpful? Share it with others.
Written by Dylan Jovine