Over my nearly 30-year career tracking the stock market professionally, I’ve seen a lot of trends come and go.

Investors who jumped on these trends late paid a stiff price in terms of lost performance, and heading into 2026, that’s been the case with the Magnificent seven stock trade. In 2025, only two of the seven actually beat the S&P 500 benchmark index.

In 2023, Bank of America analysts coined the phrase “Magnificent Seven,” invoking the famous Western to describe technology kingpins most likely to benefit from the tidal wave of AI activity unleashed by the launch of ChatGPT in 2022.

The companies comprising the Mag 7 (Alphabet, Nvidia, Tesla, Microsoft, Meta, Apple, and Amazon) quickly became “one-trade” stocks — buy the basket and watch the gains pile up. While true initially during the AI build-out, this has become increasingly less true as time has passed.

Last year, many of the stocks in the set-it-and-forget-it Mag 7 trade lagged behind the S&P 500, which rose 16.4%, surprising many.

Magnificent 7 stock returns versus the S&P 500 (2025):

Overall, only Alphabet (GOOGL), which has seen massive adoption of its AI chatbot, Gemini, and Nvidia (NVDA), which remains the AI picks-and-shovels workhorse, generated alpha, or excess return, against the benchmark.

Only two of the Magnificent 7 stocks outperformed the S&P 500 in 2025.

TheStreet/Shutterstock

AI hype shifts to “show me the money” reality

The fact that only two of the seven stocks most exposed to AI investment generated returns better than the market is more than an oddity; it’s a reflection of a maturing AI trend.

In the early days of the AI trade, winners and losers were harder to see, making a simple basket approach consisting of the Goliaths the easiest way to gain exposure.

That’s not as true anymore. The AI buildout continues at a rapid pace, but it’s getting clearer who else may benefit from AI development.

Undeniably, Nvidia deserves its seat upon the throne as the AI go-to, given that its GPUs remain the fastest and most efficient chips for training and inference, the fancy name for using AI apps after they’re developed.

Related: Investors quietly pile into a group of stocks for 2026 (it’s not tech)

However, it’s no longer the only game in town. For instance, Alphabet partnered with Broadcom (AVGO) to develop Tensor Processing Units, specially designed AI chips that can handle specific AI tasks well enough to reduce reliance solely on Nvidia.

Perhaps unsurprisingly, the growing use of TPUs made Broadcom a much better stock market performer than most of the Mag 7 last year, with its shares climbing 49%.

The AI trade also rewarded memory maker Micron (MU) and companies selling the gear necessary to connect all the new high-end server racks, such as Credo Technology (CRDO), which makes interconnect cables.

Demand for memory was so high in 2025 that spot memory prices soared, causing Micron shares to surge 239%. Meanwhile, more and larger data centers meant pricing power for Credo Technology, sending its shares 114% higher in 2025.

And we shouldn’t ignore other top AI stocks, includingPalantir (PLTR), whose platform and data analysis chops have made its software a go-to for enterprises eager to build AI agents to boost productivity. After posting 63% revenue growth and 110% earnings per share growth in the third quarter, Wall Street is modeling Palantir’s full-year 2025 revenue and EPS growth of 54% and 76%, respectively. That’s pretty magnificent.

Also Read: What’s next for Palantir stock in 2026?

The reality of the AI trade today is that it has become less about who is spending the most to build out the networks and more about who is experiencing the fastest revenue and profit growth as a result.

Sure, the Mag 7 underperformers, Tesla, Microsoft, Meta, Apple, and Amazon, have AI tailwinds. Still, they’re so big that they’d need truly massive ROI from their AI spending to generate the kind of revenue and profit growth delivered by Micron, Credo Tech, or storage provider Western Digital (WDC) last year.

Since most had already bought the Mag 7 in 2023 and 2024, there wasn’t enough money on the sidelines relative to these other underowned stocks to move the share price needle.

Will 2026 be more of the same?

It certainly seems to me that the bar has been set higher for the Mag 7 stocks, particularly the hyperscalers that are spending the most on building out their data centers. Capex at the largest hyperscalers is expected to be $527 billion in 2026, according to Goldman Sachs, up from $394 billion in 2025.

The sheer amount being spent is no longer easily paid out of cash flow, and increasingly, companies, such as Meta Platforms, are turning to the bond market to borrow money and help finance their AI plans (it raised $30 billion in a bond sale in October).

More AI Stocks:

• Morgan Stanley sets jaw-dropping Micron price target after event
• Bank of America updates Palantir stock forecast after private meeting
• Morgan Stanley drops eye-popping Broadcom price target
• Nvidia’s China chip problem isn’t what most investors think
• Bank of America sets AI stocks to buy list for 2026

The big companies receive favorable bond rates, but the money isn’t free, and as a result, the hurdle for returns necessary on those investments is higher than it was in 2024.

Yet, just as extrapolating Mag 7’s 2024 success into 2025 was mistaken, thinking Mag 7 underperformers will underperform yet again in 2026 isn’t a given, either.

“Investor willingness to endorse substantial capex upgrades for well-capitalized large AI hyperscalers will likely depend on the path to AI monetization and underlying cash flow strength,” wrote Goldman Sachs analysts in a research note shared with TheStreet.

Fortunately, these companies are increasingly better at figuring out how to make money off of AI than they were two years ago, and the appetite for AI among Main Street and businesses is high enough that they’re transitioning toward monetization.

For example, Microsoft is increasing the price of Microsoft 365 this year, citing all its AI features. Meta is already pocketing tens of billions in revenue from AI, due to leveraging its efforts for advertising upside.

Apple, which has taken its fair share of knocks for arguably being “late” to offer AI, is expected to finally make a splash in 2026 that could kick-start animal spirits. And Amazon can’t be discounted, given AI’s potential to accelerate e-commerce sales with better product targeting.

In short, there’s little reason to abandon these titans entirely, and plenty of reason to think that last year’s laggards could post better returns in 2026.

Still, it’s unlikely that those companies will be the stock market’s top performers, or that they’ll even be the best-performing stocks to ride the AI wave this year. They’re too big, and the law of large numbers is working against them.

What can investors do now?

The 2026 story may be similar to that of 2025, in that investors should approach technology stocks tactically. The set-it-and-forget-it approach may provide suitable returns for investors. Still, I suspect that the stock market’s best returns may come from unexpected, rather than expected, stocks.

For this reason, if you’re a more active investor, regularly screening for technology’s top performers every month may be a solid approach to spotting 2026’s emerging stars.

One basket I’m watching is the semi equipment manufacturers, which I think could continue rallying as more and larger semiconductor fabs are built to close the demand-supply gap and fulfill the White House’s “Made in the U.S.” game plan (I wrote more about them here).

Trading around the hyperscalers and Mag 7 names may also prove better than buying and holding, but it’s downright brutal to be right twice (selling and buying), let alone multiple times throughout the year.

Moreover, I believe it’s important to remember that although five of the seven Mag 7 stocks underperformed, owning the entire basket still beat the S&P 500. Thanks to Nvidia and Alphabet’s strength, the Roundhill Magnificent Seven ETF (MAGS) gained 23% last year.

This year? Maybe Apple or Amazon does the heavy lifting.

Todd Campbell owns shares in Apple, Microsoft, Amazon, Alphabet, Nvidia, Meta Platforms, Tesla, Broadcom, Credo Tech Group, and Micron.

Related: Every major analyst’s S&P 500 price target for 2026