AI’s Next Chapter: Profits, Discipline, and a More Durable Cycle

Let’s be honest, when you see tech stocks soaring again, it’s hard not to think about the dot-com bubble. Clients ask us about it all the time: “Is this another 1999?” It’s a fair question. The headlines are huge, valuations are up, and AI seems to be everywhere from your phone to your portfolio. But here’s the thing: this time really does look different. Not perfect, but different in some very important ways.

Solid profits, not wild guesses

Back in the late 1990s, a lot of tech companies were all promise and no profit. Today’s leaders (think chipmakers, cloud giants, and software firms) are making real money. They’re profitable, sitting on strong balance sheets, and using that cash to reinvest in their businesses. The average large tech stock trades around 30 times earnings right now. That’s not cheap, but it’s nowhere near the 55 times earnings we saw during the dot-com peak.

And those earnings are real. During the last big bubble, stock prices jumped almost 480% while earnings grew less than 80%. In today’s AI cycle, stock prices are up about 120%, roughly in line with earnings growth. In other words, investors are paying for actual performance, not just dreams of what could be.

Funded by Cash but Debt’s Creeping In

Lately, some of the biggest tech players have begun borrowing to help fund their AI investments. That’s drawn attention because these companies already generate massive cash flows. When even cash-rich firms start issuing debt, it raises the bar for what kind of returns those projects should produce. We’ve entered more of a “prove it” phase, where investors want to see tangible results, not just vision.

Still, the bigger picture matters. Even with this new wave of debt, about 80% to 90% of planned capital spending from large tech firms is still coming from their own cash. So yes, leverage is creeping in, but it’s not a red flag. It’s more about managing flexibility while keeping growth on track. Most investors remain comfortable with the trend because the core story hasn’t changed, these are still highly profitable companies with strong competitive positions and healthy balance sheets.

Real demand, real growth

AI isn’t just a buzzword; it’s changing how the economy works. Data centers (the physical backbone of AI) are expected to grow nearly 20% a year through the end of the decade. That’s huge. And it’s not just about chips. It’s about power grids, cooling systems, fiber cables: the infrastructure needed to keep these systems running.

In fact, AI-related investment added roughly 1% to U.S. GDP growth this year. That’s a real economic impact, not speculation on future profits. Companies across industries, from healthcare to manufacturing, are already using AI to save money and boost productivity.

Investors are more careful this time

During the dot-com era, everyone wanted a piece of the action. People were quitting jobs to day-trade. Money flooded into tech funds with little regard for what they actually owned.

Today’s investors look more cautious. U.S. stock funds overall have lost money this year due to outflows, while tech funds have seen only modest inflows, nothing like the frenzy of the late 1990s. That tells us people are optimistic, but still careful.

What could go wrong?

Of course, no story is risk-free. Two things could throw this off course:

1. Slower spending: If companies pull back on AI investments sooner than expected, growth could cool.
2. Too much concentration: A handful of mega-cap names still dominate the AI theme. If one stumbles, it could rattle the market.

These aren’t deal-breakers, but they’re good reasons to stay diversified and resist the temptation to chase the hottest names.