Listen to the audio version of this article (generated by AI).
One of the most profitable questions investors can ask during a massive investment boom is surprisingly simple:
What’s becoming scarce?
Our technology expert, Luke Lango, argues that the AI Revolution has been a rolling series of these shortages. First, it was advanced GPUs. Then servers, cooling, electricity, and memory. Each time, billions of dollars chased a limited supply – and enormous pricing power flowed toward the companies controlling the bottleneck.
Now, Luke believes Nvidia’s latest earnings report offers clues about where that bottleneck may be headed next.
In today’s Sunday Digest, he explains what Nvidia’s results reveal about the changing AI supply chain and why he’s applying the same thinking to Elon Musk’s sprawling empire. Specifically, Luke is asking a question that could point toward an entirely different group of AI winners: What does Musk still have to buy?
Luke will reveal what his research uncovered during his , alongside Louis Navellier and Eric Fry. They’ll map out Musk’s AI ambitions, the critical bottlenecks they see developing, and the suppliers that could benefit – including one ticker Luke will reveal free during the event. .
Before we hand it over to Luke, a quick note…
This Tuesday at 11 a.m. ET, veteran trader Jonathan Rose is hosting legendary investor Louis Navellier on Jonathan’s free episode of .
Louis will dive into AI’s real choke point today, where he’s putting money to work in energy, and why the S&P’s 53% earnings quarter marks a near-term peak.
, and we’ll see you Tuesday.
Now, back to Luke…
Have a good weekend,
Jeff Remsburg
On a plateau south of Bandung, on the island of Java in what is now Indonesia, Dutch planters once cultivated one of the most strategically important plants on Earth.
It was called cinchona.
Its bark contains quinine, which was then the world’s best defense against malaria.
If European powers wanted to build railroads, man military outposts, or expand deeper into the tropics, they needed quinine. And the Dutch controlled almost all of it.
By the 1920s, plantations in the Dutch East Indies supplied more than 90% of the world’s cinchona. For years, the rest of the world paid the toll.
Credit: iStock/Nastasic
Cinchona bark was once the world’s primary source of quinine, and control of its supply gave the Dutch enormous leverage.
Then World War II came. Germany occupied the Netherlands in 1940. Japan captured Java in 1942.
Almost overnight, the Allies lost access to the cinchona plantations and much of the infrastructure needed to turn their bark into medicine.
The consequences were brutal. Malaria ravaged troops fighting in the Pacific. Thousands of American soldiers were hospitalized with the disease as military planners scrambled to secure alternative supplies and treatments.
A massive military operation had been built around one relatively small, scarce resource. And almost nobody appreciated just how important it was until the supply disappeared.
Today, we’d call quinine a bottleneck: one relatively small piece of a much larger system that can determine how well the entire system functions.
Bottlenecks have an interesting economic consequence. When everybody desperately needs the same scarce thing, the people who control it gain enormous leverage.
That’s the idea I want to talk about today.
While we often talk about it as one trade, the AI Revolution really has been a rolling series of bottlenecks.
And the latest earnings report from Nvidia Corp. (NVDA) just gave us another clue about where the next one may be forming.
Follow the Bottleneck
First, Nvidia’s August 26 report helped us understand just how powerful AI demand remains.
In it, Nvidia gave investors unusually clear guidance about its next fiscal year. Management expects revenue growth of around 70% – far above what Wall Street had been anticipating.
For a company already operating at Nvidia’s enormous scale, that’s remarkable.
But another number caught my attention even more.
Nvidia expects its gross margins to decline.
In plain English, that means Nvidia expects to keep a little less profit from each dollar of product it sells.
Normally, that data point would suggest trouble. But here, it tells us something very different.
Nvidia has plenty of customers with lots of money to spend. Demand is so strong that Nvidia says it could sell even more GPUs if it could secure enough of the components it needs.
And that means Nvidia is having to pay up for scarce supplies. It’s run into a bottleneck of its own.
Some of the pricing power that once belonged almost entirely to Nvidia is moving further down the supply chain – toward the companies making the things Nvidia itself desperately needs.
We’ve watched this happen repeatedly throughout the AI boom.
First, there weren’t enough powerful computing chips, and Nvidia exploded (it was the bottleneck). Then there weren’t enough servers to house those chips, cooling systems to keep them from overheating, or electricity to power the enormous data centers being built around them.
More recently, memory became scarce. AI systems need enormous amounts of specialized memory to rapidly feed data into their processors. That shortage has helped companies like Micron Technology Inc. (MU) command higher prices.
Each time, the same basic thing happened.
Source: ChatGPT image
A huge amount of money rushed into AI. The buildout ran into a constraint. Whatever was scarce became more valuable.
And a new group of companies got paid.
Now the bottleneck appears to be moving again.
The Next Constraint
This time, the problem is networking.
Here’s why that matters.
Imagine putting 100,000 of the world’s fastest computer chips inside a data center.
Having all that computing power doesn’t do much good if those chips can’t communicate with one another fast enough. They need to work together almost like one enormous computer.
Increasingly, the connections carrying all that information use light to move enormous quantities of data at extraordinary speeds. Suddenly, the equipment connecting the chips has become almost as strategically important as the chips themselves.
Nvidia is showing us how important this is becoming. Earlier this year, the company committed billions of dollars to secure supplies from optical-component makers including Lumentum Holdings Inc. (LITE) and Coherent Corp. (COHR).
Companies make commitments like that when they’re worried they won’t be able to get enough.
In other words, Nvidia is showing us where one of its own bottlenecks lies. That’s exactly where investors want to look for opportunities.
And Nvidia’s supply chain isn’t the only place I’ve been looking.
What Does Elon Still Have to Buy?
For months, I’ve been applying this same idea to Elon Musk.
That’s because Musk hates bottlenecks. When something his companies need is too expensive, too scarce, or too slow to arrive, his instinct is to figure out whether he can build it himself.
Tesla Inc. (TSLA) has brought important parts of automobile manufacturing in-house.
Space Exploration Technologies Corp. (SPCX) builds its own rockets and engines while operating the Starlink satellite network.
And xAI has built enormous computing clusters to train and operate its artificial intelligence systems.
Put all of Musk’s businesses together and you begin to see just how much of the AI infrastructure stack he increasingly controls himself.
That raises a potentially profitable question for investors: What does Elon Musk still have to buy?
After all, if Musk has spent years trying to eliminate expensive or unreliable suppliers, the companies that remain indispensable to him deserve a very close look.
Whatever the reason, Musk still has to write those firms checks.
And when you’re talking about the enormous ambitions of Tesla, SpaceX, and xAI, those can become some very large checks.
I’ve spent much of this year studying Musk’s companies, the technologies they need, and the suppliers that could become increasingly important as his AI ambitions grow.
And I’ve identified a small group of companies that I believe could sit at some of the most important bottlenecks in Musk’s next phase.
That’s what I’m going to show you in just a few days.
At 8 p.m. ET on Wednesday, September 9, I’m hosting a free InvestorPlace workshop devoted to what I believe Elon Musk is building next – and, more importantly for investors, the companies he may have to pay to build it.
I’ll be joined by my colleagues Louis Navellier and Eric Fry.
Together, we’ll walk you through the research, show you where we believe some of Musk’s most important bottlenecks are developing, and explain why we think a handful of little-known suppliers could be positioned to benefit. I’ll even give you the ticker of one company we’re watching completely free.
And there’s another reason I want to do this now.
I’ve circled the date September 24 after studying Musk’s plans and recent developments surrounding his companies. Nothing involving Elon Musk is guaranteed, of course. But I believe something could happen around that date that removes an important obstacle standing between Musk and the next phase of his AI ambitions.
Which is why I want you prepared before then.
You can . It’s completely free, and registration takes only a few seconds.
For now, remember the lesson Nvidia just gave us: Follow the bottleneck.
When something becomes scarce in the middle of a massive investment boom, power shifts toward whoever controls it.
That was true when armies depended on quinine from trees on Java.
It was true when Nvidia’s GPUs became the indispensable engines of the AI boom.
And it’s happening again as Nvidia itself scrambles to secure the components it needs.
That’s why you need to know what Elon Musk still has to buy. That’s where some of the AI Revolution’s next great investment opportunities are hiding.
Sincerely,
Luke Lango
Senior Investment Analyst, InvestorPlace
P.S. Over the years, Luke has recommended 33 stocks connected in one way or another to Elon Musk’s businesses that went on to double or better at their highs. Now he’s mapped Musk’s expanding AI empire to identify the suppliers that he may need for what comes next. , and he’ll show you what he found – including one ticker you can take away from the event completely free.
