For one day in September, Larry Ellison briefly became the richest man in the world.

It was thanks to a soaring stock price at Oracle, the software and cloud-infrastructure company he founded in 1977.

Oracle had just reported earnings. The numbers were strong, but it was the company’s projections that stunned Wall Street.

Due to rising demand from AI, Oracle said its cloud-infrastructure business could grow from about $10 billion a year to $144 billion by 2030—a 14x leap.

Oracle’s share price rocketed nearly 40% in one day on the news. And the company became the new poster child for the AI boom.

Three months later, the tone has changed.

Last week, Oracle reported another solid quarter. Earnings beat expectations. Revenue grew 14%. Big AI customers—including Meta, Nvidia, OpenAI—are still signing massive deals.

But this time after its earnings results, Oracle’s share price fell. It’s now down about 45% from the high it set after its September surge.

If the AI boom was red hot in September, it’s closer to lukewarm now.

That doesn’t mean the boom is over. But investors are starting to ask harder questions about the cost of the infrastructure required to sustain it… and repricing stocks that could struggle to foot the bill.

And as we’ll explore today, Oracle may be the canary in the coal mine for other popular AI stocks.

As I’ve been hammering on in these pages over the past couple of months, now is the time to put a risk management plan in place. You want to make sure the collapse of the AI trade—regardless of when that happens—doesn’t become an extinction-level event for your wealth.

That’s what Tuesday’s Tipping Point 2026 event was all about. Along with “quant” investing legend Marc Chaikin, we’re circling the wagons to make sure you protect yourself and prepare.

And we launched a groundbreaking new technology to help do just that. We designed it with volatility and bear markets in mind to help you avoid the fast, whiplash selloffs they can produce.

If you’re one of the more than 10,000 people who joined us on Tuesday, thank you! I hope you walked away with a clear plan for how to deal with the kind of volatility Marc and I see coming.

Today, let’s take a closer look at what’s going on at Oracle… and what it’s signaling about the health of the AI trade.

From Red Hot to Lukewarm
At first glance, Oracle’s latest earnings report looked okay.

Earnings per share—a key measure of profitability—came in ahead of estimates. Revenue narrowly missed, but still grew at a healthy double-digit rate. And the company’s backlog of future AI business continued to swell.

Oracle’s remaining performance obligations—revenue already under contract but not yet booked—rose to $523 billion, up 15% in one quarter. New commitments came from some of the biggest names in AI, including Meta and Nvidia.

That’s real demand… lots of it. So why did the stock get smashed?

Investors stopped looking at Oracle’s orders and started looking at its spending bill.

To meet its AI commitments, Oracle now expects about $50 billion in capital spending this year. That’s up from an already eye-popping $35 billion estimate just months ago.

Last year, Oracle generated roughly $12 billion in free cash flow. So it doesn’t have enough internal cash flows to fund that buildout.

It’s already borrowed $38 billion and is lining up another $18 billion bond offering. And analysts now say it may need close to $100 billion in total financing to fulfill its AI promises.

That’s why, earlier this month, a key measure of Oracle’s credit risk—its five-year credit default swap spread—spiked to levels last seen during the 2008 global financial crisis. Lenders are getting nervous. And it’s souring the mood on Wall Street.

This kind of AI spending bill isn’t unique to Oracle. A report from investment bank UBS revealed that AI data center and project financing deals have surged to $125 billion this year from $15 billion in the same period in 2024. ​And even more debt is needed again next year.

Maybe this is why other giant AI stocks have been selling off recently, too. For example:

  • Nvidia is down about 15% from its 2025 peak
  • And Google parent Alphabet is down about 12% from its recent high
  • Microsoft is down about 10% from its peak

These aren’t massive moves. But together, these three mega-cap AI stocks account for about 19% of the entire S&P 500 market cap. If they continue to sag, the rest of the market won’t be far behind.

I’m not an AI doom-and-gloomer. I have some concerns about how this technology will impact society, especially the labor market. But I’m also excited about AI’s potential. My team and I have even built a suite of AI-powered tools to help spot short-term trade setups in the market a human trader might miss.

But whatever your views on AI, it’s vital you protect your downside from sudden drops. Whenever a trade gets this crowded, there’s a risk that it will suddenly reverse.

And the crowded nature of the AI trade isn’t the only concern as we head into 2026.

As Marc established during Tuesday’s event, there are also the bearish implications of the four-year presidential cycle to consider.

Surprisingly Consistent Rhythm
Nearly all of the top brokerage houses—including T. Rowe Price, Morgan Stanley, Charlies Schwab, and Goldman Sachs—have studied this cycle.

Even super investor Warren Buffett considers it when making his investments.

It’s easy to see why…

For more than a century, U.S. stocks have followed a surprisingly consistent rhythm tied to the four-year presidential election cycle. It’s not about politics—the cycle tends to hold no matter who’s in power. It’s about how markets react to changes in government policy.

Markets tend to boom in the run-up to elections, when administrations lean hard on stimulus, growth, and giveaways.

Once the votes are counted, that largesse fades, and reality sets in.

In a nutshell, the four-year cycle breaks down like this:

  • Pre-election and election years are usually the strongest for stocks
  • Post-election years are where volatility creeps back in
  • Midterm years are where markets most often stumble

The post-election year—for the current cycle, that’s 2025—is where we first start to see the cracks forming.

The worst year tends to be the midterm election year. In this case, 2026.

Across the last 14 election cycles, bear markets began or were in progress during nine of those midterm years. In other words, we saw a bear market about two-thirds of the time.

Which is why Marc is warning that 2026 could be the Year of the Bear.

Sincerely,

Keith Kaplan

Get This Stock Now [sponsor]
This company is the lifeblood of AI data centers, yet almost no one has caught up with the story. Their hardware is so essential that the data center industry uses enough of it to stretch around the world 8 times – in a single building! So, if you own Nvidia stock now, you might be well-served to sell those shares and check out this under-the-radar play instead. Or if you missed the boat on Nvidia, this is a rare second chance to target tremendous profit potential as AI data centers spring up in every corner of the world. Get my full take on this exciting play right here...

Source: TradeSmith