Listen to the audio version of this article (generated by AI).
Editor’s Note: Joe Austin has spent four decades on Wall Street — as a tech-sector research analyst, a senior portfolio manager overseeing more than $10 billion in assets, and a coverage analyst for a $5 billion hedge fund. In that time, he’s learned that the biggest profits in any major technology wave rarely go to the most obvious players.
In today’s piece, Joe uses an unforgettable moment from NASCAR to explain why the most overlooked AI stocks may be the most valuable ones — and why the “behind the scenes” companies making AI physically possible deserve a serious look.
Joe recently partnered with 60-year Wall Street legend Marc Chaikin to debut a first-of-its-kind AI-powered tool. It’s the first AI-powered product Chaikin Analytics has ever built, designed specifically to find stocks like the ones Joe describes below, before Wall Street catches on.
Just yesterday, Joe and Marc held a special presentation detailing the tool. You can catch a replay of that here.
Now, here’s Joe…
Denny Hamlin was ready to clinch his first NASCAR championship. But it all fell apart because of a piece of tape.
Hamlin was considered one of the best drivers of his generation, but the title had always eluded him.
In 2019, he was NASCAR’s comeback kid. He’d already won six times that season — including the iconic Daytona 500. The championship all came down to a cool November night at the Homestead-Miami Speedway.
The day started well. Qualifying had been canceled due to bad weather, which gave Hamlin the pole position based on his points standing. In the first two stages, he finished fifth. By the third stage, things were looking up — he was running second, posting quicker lap times than his teammate Kyle Busch, who led the race.
Then, with 58 laps to go, crew chief Chris Gabehart made a fateful call. He brought Hamlin in for an early pit stop and had the crew slap a big piece of black tape on the front grille.
The idea was sound, in theory. In NASCAR, crews routinely tape grilles to gain a competitive edge. Normally, air enters the grille and bounces around the engine, creating drag. Tape restricts that airflow, forcing it to flow smoothly over the car instead. More speed, more downforce, better grip.
But tape also restricts cooling.
NASCAR engines run at around 290 degrees Fahrenheit — about 90 degrees hotter than a typical road car. The margin for error is razor-thin. In Hamlin’s case, the tape backfired immediately. Temperature gauges maxed out. Steam started spewing from the engine. Engine failure seemed imminent.
Gabehart had to call Hamlin back to the pit after just 12 laps. He finished 10th — dead last among the four championship contenders. Kyle Busch won both the race and the title.
Excessive heat isn’t just a NASCAR problem. Data centers running AI chips face the exact same dilemma.
And for investors, it represents one of the most overlooked opportunities in the entire AI boom.
Finding the “Behind the Scenes” Companies
AI chips need massive amounts of power to train models and run computations. More power means more heat. And if you can’t cool the chips fast enough, performance crashes, or the hardware fails entirely.
Inside a data center are long rows of computer racks — tall cabinets stacked with servers. A typical AI data center contains hundreds or even thousands of them. Nvidia Corp.’s (NVDA) next-generation Vera Rubin chip uses 120 to 130 kilowatts per rack. That’s the annual electricity consumption of about 100 U.S. homes — per rack.
Bigger versions of the Rubin chip will use five times that much power.
That creates an unavoidable physics problem. More power means more heat, in a nearly one-to-one relationship. And delivering this much electricity requires completely rethinking how power gets moved through a building.
Think of it like water through a hose. You can deliver the same volume using high pressure through a small hose or low pressure through a massive one. Electrical power works the same way — high voltage with low current, or low voltage with high current. High current dangerously overheats cables. Traditional power systems can’t handle it.
Engineers solved this by raising the voltage. The industry has shifted to 800-volt power systems, which deliver the same power with far less current and far less heat.
But operating at 800 volts requires power chips made from entirely different materials. Materials that only a handful of companies know how to produce.
This is the bigger point about investing in a megatrend like AI.
The AI technology itself gets all the media attention. Nvidia, Microsoft Corp. (MSFT), Palantir Technologies Inc. (PLTR): Everyone knows those names.
But an entirely different set of companies — the ones making the components, materials, and systems that let AI physically function — are just as essential. And far less picked over.
No matter which big-name AI company is building the next data center, the “behind the scenes” businesses making it all run are going to get paid. The question is whether you own any of them.
A Tool Built for Exactly This
I’ve spent 40 years on Wall Street learning to look one step behind the obvious story. The internet boom made millionaires out of people who bought Cisco Systems Inc. (CSCO) and Intel Corp. (INTC), not just Amazon.com Inc. (AMZN) and eBay Inc. (EBAY). The shale revolution enriched investors in fracking equipment and pipeline infrastructure, not just oil producers.
The AI boom is setting up the same way. And the challenge — as always — is finding the right “behind the scenes” stocks before the crowd does.
That’s exactly what Marc Chaikin and I built the Time Machine to do.
The Time Machine is Chaikin Analytics’ first-ever AI-powered platform, and we unveiled it for the first time yesterday during a special free broadcast (you can watch a replay here). It works by scanning decades of market history to find stocks today whose fundamental and technical fingerprints match the early profiles of proven multi-bagger winners — stocks like Nvidia, Amazon, and Meta Platforms Inc. (META), right before their biggest moves.
In backtesting, it surfaced stocks that went on to deliver gains of 995%, 1,406%, and 3,804% — all while the “seed” stocks they were matched against posted far more modest returns.
The picks-and-shovels AI companies — the power chip makers, the cooling system specialists, the infrastructure suppliers — are exactly the kind of stocks the Time Machine is designed to surface.
This is the first time Chaikin Analytics has ever offered an AI-powered product.
Click here to watch our newly released special broadcast and learn more about Time Machine.
Good investing,
Joe Austin
Senior Analyst, Chaikin Analytics
