MW Think AI is a bubble? Here's what you can do about it.

By Christine Ji

The 'easy money' in the AI trade has already been made. These three strategies can help you protect your portfolio and reduce risk.

The AI boom has been fueled by hundreds of billions in capital spending from Big Tech, leading investors to question its sustainability.

Artificial intelligence has been the biggest force driving the stock market to new highs - but some investors are questioning if the rally is heading into bubble territory.

The "Magnificent Seven" megacap tech names have accounted for nearly half of the S&P 500 index's SPX gains from its April lows, signaling a highly concentrated market. Directly or indirectly, investors' portfolios are increasingly dependent on the fate of the AI trade.

"A correction in this phase will be violent," Ted Mortonson, managing director at Baird, told MarketWatch. "Everybody's in the pool."

Big Tech leaders don't seem too concerned: Meta Platforms Inc. (META) Chief Executive Mark Zuckerberg said on the Access podcast last month that the risk of being too slow to build AI was bigger than the risk of "misspending a couple of hundred billion."

But unlike Zuckerberg, the average investor doesn't have that luxury. So what's the strategy? Here's what the experts are saying.

Reduce your Big Tech exposure

Investors can lower their portfolio risk by reducing their weightings of big AI winners, according to Mortonson. As a rule of thumb, Mortonson trims 15% of a holding in a stock if its price moves below the 50-day moving average, as the negative price action can signal the start of a selloff. Taking profits can also help investors lock in gains from shares that have already run up significantly.

To the credit of Big Tech companies, they've "delivered the goods on earnings growth" thus far, Mike Reynolds, vice president of investment strategy at Glenmede, told MarketWatch. But the next leg of the AI trade will depend on their massive capital expenditures paying off - and quickly.

"Valuations aren't pricing in a slow-adoption story here," Reynolds noted. "They're pricing in something closer to a one- to two-year buildout and immediate return on investment" - something which is far from guaranteed.

Many AI-infrastructure companies, such as CoreWeave Inc. (CRWV), rely heavily on a few Big Tech customers. If the AI capex cycle peaks, it would create a domino effect - hitting not only the tech giants, but also the entire "picks and shovels" side of the AI trade.

Read: This is the critical detail that could unravel the AI trade: Nobody is paying for it.

Follow the free cash flow

While AI has the potential to be transformative for many companies, make sure your investments remain grounded in fundamentals, not speculative future outcomes. For this reason, Mortonson recommends selecting companies with robust free cash flow as a defense, in case AI revenues don't materialize.

Companies like Meta, Amazon.com Inc. (AMZN) and Google parent Alphabet Inc. $(GOOGL)$ $(GOOG)$ are seeing their free-cash-flow lines take a hit from high capital spending on AI infrastructure. But there are still opportunities in technology names, such as beaten-down enterprise-software stocks and select semiconductor companies.

Mortonson identified Qualcomm Inc. $(QCOM)$ as an overlooked AI opportunity. The company is most known for designing the processors in Android devices, and it "free cash flows very well because it's fabless," Mortonson said - referring to how the company designs its own chips, but outsources the expensive manufacturing process. Qualcomm has also moved beyond just chips, and toward automotive and industrial business lines that should benefit in the long term if AI becomes integrated into more devices.

Mortonson is also setting up his clients in software names like Salesforce Inc. (CRM) and Workday Inc. (WDAY) While both Salesforce and Workday have been dubbed "AI losers," Mortonson argues that investors are overlooking these companies' ability to generate free cash flow separate from the larger AI capex cycle.

See more: These AI 'loser' stocks were left for dead. Now it might be their turn to rally.

Consider small caps

Investors should also consider expanding their holdings beyond large-cap names. U.S. large-cap growth valuations are creeping toward the 90th percentile relative to historical averages, Reynolds pointed out, and small-cap stocks offer a rare pocket of discounted valuations.

Small-cap stocks are well positioned to outperform as the Federal Reserve begins to lower interest rates, a move that would reduce their borrowing costs and enhance the appeal of their domestic focus, Reynolds said.

Fundamentals in the space are also improving. According to a Bank of America report Friday, constituents of the Russell 2000 index RUT are increasingly seeing upward earnings revisions, especially in the last month. Compared to large caps and midcap stocks, small caps are trading at the lowest premium to historic valuations, offering investors a more attractive entry point.

A potential AI bubble could take a while to pop, Mortonson believes, as next year's AI capex estimates are still trending higher. In his opinion, the real financial consequences of this spending spree will become more prominent in 2027 and beyond, when depreciation costs for new data centers start to ramp up.

Given these trends, investors willing to take more risk may let their top-performing AI stocks ride for a little bit longer. But as more risks emerge, the long-term winners are becoming less obvious.

"The easy money has been made," Mortonson said of the AI trade.

Also read: The AI trade increasingly hinges on OpenAI - and that's a big risk for the entire market

-Christine Ji

This content was created by MarketWatch, which is operated by Dow Jones & Co. MarketWatch is published independently from Dow Jones Newswires and The Wall Street Journal.

(END) Dow Jones Newswires

October 04, 2025 08:00 ET (12:00 GMT)

Copyright (c) 2025 Dow Jones & Company, Inc.