It’s hard not to at least acknowledge this tweet from Elon Musk:
The election tomorrow has Big Tech on edge.
And for good reason.
For the past several years, they’ve had their run of the capital provided by the Western financial system.
Nvidia (NVDA) made the first fortune selling shovels for the AI boom.
The next fortunes may come from selling Nvidia’s customers everything else they forgot they needed.
Start with electricity.
Goldman Sachs expects U.S. data-center power demand to rise 38% this year and another 38% in 2027. The bank recently raised its estimate of U.S. data-center capacity at the end of 2026 to 64 gigawatts and expects 90 gigawatts by the end of next year.
The technology promising to make everything more efficient is helping drive a surge in energy demand that requires a massive buildout of its own. (Source: Precedence Research)
Getting that electricity to the computers has become its own industry.
Nexans CFO Vincent Piquet told Reuters last week that cable manufacturing capacity is becoming another bottleneck. A new production line for the high-voltage cable used by power networks can take roughly three years to build. Grid constraints are already delaying some data-center projects.
Goldman now calls power the AI industry’s enduring bottleneck. The bank estimates global data-center electricity demand will rise 170% between 2025 and 2030 and calculates roughly $7.6 trillion of spending across computing, data centers and power infrastructure through 2031.
But $7.6 trillion of spending is not $7.6 trillion of wealth creation. Spending measures what goes into the boom. The harder question is what economically productive assets come out the other side — and what return those assets earn after depreciation, power costs and financing.
Our first trade follows that bottleneck into a part of the semiconductor business most investors never see.
Our second follows the Pentagon.
Artificial intelligence has moved from Silicon Valley demonstrations into military operations. Palantir’s (PLTR) Maven Smart System now processes information from satellites, radar and other sensors to help military officers assemble a battlefield picture faster. NATO bought Maven for its Supreme Headquarters Allied Powers Europe in 2025 and declared it fully technically operational this June.
That changes the character of the AI trade.
Governments buy differently from consumers. Contracts can run for years. Security requirements create barriers to entry. Once software becomes embedded in command systems, replacing it becomes more complicated than cancelling a streaming subscription.
Our second trade follows that stream of government and military spending.
The third follows the enormous private capital commitments required to keep the whole machine running.
Reuters reported this month that S&P Global believes the largest technology companies may carry more AI-related liabilities than investors can readily see on their conventional balance sheets. Rising rates have already made investors more discriminating about AI debt. CoreWeave’s bonds and shares came under pressure this month as investors questioned the financing behind its rapid expansion.
That’s where the shovel analogy becomes especially useful. The best position in a capital boom is often held by the company that gets paid whether the ultimate project earns an adequate return or not. The buyer carries the risk that the new capacity disappoints. The supplier receives payment for the invoice today.
The companies we’re eyeing don’t need to know which chatbot or tech giant ultimately wins the race. They just need the race to keep going. And with Big Tech writing enormous checks to build out its AI infrastructure, the companies selling the tools behind the revolution could deliver huge returns. (Source: Shutterstock)
The AI boom is growing up.
The first phase rewarded companies that could promise more computing power.
The next phase will reward companies that can solve specific physical and institutional problems: electricity, power management, specialized chips, military deployment and federal procurement.
Those are businesses with invoices attached.
Our system is signaling three trades along those fault lines heading into Nov. 3. Andrew has taken the signals apart, studied the companies and built the options structures we intend to use.
The election gives us a date.
The AI buildout gives us the spending.
The bottlenecks tell us where to look.
~ Addison
P.S. Tomorrow, one of two things is going to happen. Republicans will retain control of Congress and the Trump mandate will kick into high gear for the remaining two years of his second term.
Or, the data proves out, and socialists move into positions to stall tax reform, impeach administration officials they don’t like and let TDS run wild through the mainstream press all over again.
Either way, it’s up to you to keep your own eyes on your money, right?
Take a look at this:
That’s a photo we just took of the actual patent for our AI-driven forecasting tool. It’s remarkably accurate. The AI tool sifts through millions of transactions in the open market simultaneously and gives us strong forward guidance on any stock we choose to run through the system.
It’s fantastic. And gives us insight into larger-than-life macro trends as well as granular price movements on individual stocks.
Why does the patent matter?
Well, the good news is you don’t have to forecast political winners or agree with the flaccid promises candidates make to get elected to find potential winners on Wall Street.
The patented tool also helped us assemble the exact historical patterns that will help you make and preserve your money before, during and after the midterm elections and the investment opportunities that will naturally flow from a political shift in Washington. Click here to see how you could put them to work in your portfolio.
