Massive AI Bets, Slowing Economy Could Lead to Stock Market Crash

Wall Street really needs AI to live up to the hype.

A lot has been said about the emerging technology’s world-changing potential: Its ability to create stunningly realistic images and videos, ace the LSAT and the MCAT, and complete rote research tasks. You could argue it’s ready to augment — or even replace — entry-level jobs.

These features have investors up and down the Street very excited. Staunch supporters like Fundstrat’s Tom Lee and Wedbush’s Dan Ives say AI could revolutionize the human experience. Research desks from big banks like Goldman Sachs and Bank of America have given subtler nods to the prospect of AI as a productivity and profit booster, which could provide an undercurrent to stock market success over the next several years.

In fact, analysts are counting on the AI mania to fuel the market even as the White House’s chaotic trade policy eats into corporate America’s profit potential. Earnings for S&P 500 companies are projected to grow 8% this year, a fairly average showing for an anything-but-average year. What is notable is just how much of that growth relies on the tech sector: Silicon Valley companies are expected to boost their earnings by 21% — the highest growth of any sector. By contrast, profits for retailers are forecast to grow a measly 2.5%.

Within the tech sector, semiconductor companies — one of the most globally exposed industries on the stock market — are expected to supercharge profit this year, with a projected climb of 49%. This enthusiasm is a signal Wall Street is betting that demand for AI’s use cases will supersede tariff turmoil or job market wobbles.

AI’s growth has been incredible, and its adoption has been strong enough to leave its fingerprints on economic data sets like business investment and manufacturing spending. Yet no matter how rabid the world is about AI’s possibilities, the amount that investors are relying on the tech to fuel the market’s gains — especially in the face of rising economic uncertainty — feels short-sighted. Tech stocks helped the market recover from its April malaise, yet earnings expectations and economic momentum are even weaker than they were at the lowest point of the sell-off. This combination leaves the stock market in a precarious spot: Either AI needs to live up to the hype, or investors could be looking at a gnarly second half of the year.

One way to tell the story of human history is through our technology — the lightbulb, the calculator, the tractor, the computer all stand as markers to our societal progress and have helped drive the level of efficiency and productivity we enjoy.

Tech has perhaps played an equally prominent role in our investment portfolios. The Magnificent Seven stocks — Apple, Amazon, Alphabet, Meta, Microsoft, Tesla, and Nvidia — are collectively worth $18 trillion, or about 33% of the S&P 500’s total market value. Together, their stock prices have increased 330% over the past five years, compared with a 100% rise in the S&P 500.

It makes sense. Big Tech’s products have become deeply integrated into our daily lives, and that level of ubiquity has also captured Wall Street’s attention. Venture capital fundraising reached a record in the first quarter amid a huge appetite for AI investment, and S&P 500 companies mentioned AI more than tariffs on second-quarter conference calls.

The $65 trillion US stock market may be particularly gripped by Big Tech’s ups and downs these days, but it hasn’t always been this way. Tech has averaged about 20% of the S&P 500’s market value over the past decade, including 13% in the five years before COVID. The dominance is not set in stone, and while the wider market’s fortunes are tied to tech now, that may not always be the case.

While the stock market may seem like one big proxy for the tech sector’s explosive growth right now, there is one deeper connection that should draw investors’ attention. Over time, the S&P 500 has been attached at the hip to the fate of the broader US economy. Eight of the past 12 market crashes — S&P drops of 20% or more — overlapped with recessions. No matter how high-flying an industry is, recessions tend to pull stock prices and business hopes back to Earth. The internet revolutionized the world in the late 1990s, and the explosion in social media dominated the 2010s, but the information sector has shed employees and seen share prices fall in the past three recessions.

Given that setup, we’ve set the stage for a portfolio smackdown of the ages. Economists are worried a recession is coming, yet investors are surprisingly upbeat about AI’s prospects — so upbeat that they’ve bid S&P 500 tech stocks to nearly a record high. Sell-side analysts who evaluate company-level trends are similarly optimistic. But in the real economy, layoffs are growing, and hiring has ground to a halt. The sharply diverging views between economists and stock analysts mean someone has to be wrong. The freight train that is AI adoption — a three-year story of rapid innovation and progress — could collide with a massive wall from historic tariffs, high interest rates, and low consumer confidence.

What’s particularly rich about this is that tech companies are the most exposed sector to global tariffs. They gather the highest percentage of revenue internationally, plus they have the most suppliers and factories outside US borders. In fact, semiconductor companies — the firms providing chips for AI technology — are expected to hit that aforementioned 49% earnings growth despite generating 67% of their revenue abroad and sourcing 70% of their supplies from overseas.

Some analysts believe that if AI hopes can keep the stock market chugging along, maybe it can do the same to the economy. After all, companies invested an inflation-adjusted $2.2 trillion on computers and other processing equipment last quarter, about one-seventh of the $16 trillion Americans spent on goods and services. Investing more in AI does ultimately help boost the economy, but that $2 trillion is peanuts compared with the real engine of the US economy. Americans’ spending accounts for about 70% of GDP — by far the biggest driver of output — and spending has dropped in each of the past nine recessions. If tariffs intimidate consumers and lead to layoffs that decimate American incomes, then the economy is probably bound for a crisis — whether the robots pan out or not. And based on history, an economic crisis could topple the stock market.

The math shows us that AI isn’t much of a match for some effects of tariffs and may not logistically be enough to save the economy from ruin. Your portfolio’s outcome may be a different story, though.

This is when I have to introduce you to one of the most frustrating adages of investing: The stock market is not the economy.

The economy is the value that we create — the hard assets, the cash spent, the paychecks we get. Stocks are an expression of that value, but they use the present reality to project future expectations. AI’s impact on the economy may not bear out through numbers. But in your portfolio, AI’s influence depends on how willing we are to collectively dream up better days ahead in terms of what AI is capable of and how much money AI-dominant firms will amass in the years to come. Dreaming is already a big part of the AI trade. S&P 500 tech companies’ estimated earnings grew about 50% in 2023 and 2024, yet their share prices jumped 112%.

Nothing is cut-and-dried when it comes to the stock market. It is the ultimate tangled web of logic, psychology, and mixed incentives. The stock market’s future depends on investors’ ability to dream, and people are willing to dream when they feel confident in the present moment.

The problem is, investors are awfully confident about tech stocks right now. S&P 500 tech companies made up about 23% of total index profits in the first quarter, yet their shares account for 32% of the S&P 500’s value. To close that gap, tech profits would have to grow 40%, or tech stocks would have to drop 29% from their end-of-June levels.

Stocks can thrive when expectations are higher than reality, but in these conditions, they require reasons to stay hopeful. The problem arises when investors aren’t willing to dream. When they’re too focused on present issues to give compelling stories the benefit of the doubt. Or in big market drops, crushed by financial strain.

Then, the numbers matter. People claw for any concrete evidence of AI’s value. They demand proof of profits, even though companies are spending money on a pivot to the next big thing. Stock prices adjust, and if you hold a swath of US stocks or index funds, your money is probably heavily exposed to this reality check.

This is what happened in 2000. Investors were willing to dream about this brave new technology called the World Wide Web until interest rates climbed too high and the reality of how much computing was needed for Y2K was found to be way overblown. Suddenly, the dream died, and tech stock prices came back down to reality. These days, we all know that dream wasn’t completely off base. Yet share prices took an 80% crash before the promise of the new tech came to fruition.

This is what I worry about the most in the clash between AI and the economy. We’re somewhere between AI saving the world and being an overhyped bust of a technology that can be ripped off by another country. I’m not foolish enough to call this a bubble, and I think AI will eventually deliver benefits for our economy.

We’re not there yet, though, even though investors like to think so. It takes years for big technological trends to take hold, and productivity usually shines through when workers feel empowered and companies feel comfortable expanding. That’s far from the case right now — business confidence is in the dumps, so we’re in the opposite scenario.

When the economy is getting weaker, it’s best to grasp onto what’s real in your portfolio. And there’s a striking gap between AI and reality.

Callie Cox is the chief market strategist at Ritholtz Wealth Management and the author of OptimistiCallie, a newsletter of Wall Street-quality research for everyday investors. You can view Ritholtz’s disclosures here.