Magnificent 7 is passe. This group of AI stocks can replace it

Sign up now: Get ST's newsletters delivered to your inbox

Oracle shares are up more than 75 per cent in 2025 as its AI-related cloud-computing business takes off.

Oracle shares are up more than 75 per cent in 2025 as its AI-related cloud-computing business takes off.

PHOTO: AFP

Follow topic:

- Wall Street’s most powerful collection of stocks, the Magnificent Seven, is looking a tad dated. Make way for the Great Eight. Or maybe the Golden Dozen. Or the TenAI of GenAI.

It has been almost three years since OpenAI’s ChatGPT made artificial intelligence (AI) the centre of the global economy, and in that time, one trade has dominated the US stock market: Buy the Mag Seven. Comprised of Nvidia, Microsoft, Apple, Alphabet, Amazon.com, Meta Platforms and Tesla, this septet was seen as best positioned to deliver huge returns to investors during the biggest technological shift since the internet.

While that has largely played out, a funny thing happened on the way to global dominance. The AI trade expanded in unexpected ways and moved beyond several of the market’s favourite big tech companies.

So investment strategies based on the Magnificent Seven – which is responsible for more than half the S&P 500 Index’s 70 per cent-plus rise since the start of 2023 – are missing some of the firms that are also expected to thrive in an AI future, like Broadcom, Oracle and Palantir Technologies.

“Just because the Mag Seven won past tech cycles like mobile, the internet and e-commerce, that doesn’t mean they’ll win here,” said Mr Chris Smith, portfolio manager of Artisan Partners’ Antero Peak Group. “The next winners will be the ones that address large and unconstrained markets through AI, becoming bigger companies in the future than the Mag Seven are today.”

This is not to say that the original seven are going away. The Mag Seven accounts for almost 35 per cent of the S&P 500, and its earnings are expected to climb more than 15 per cent in 2026 on the back of 13 per cent revenue growth, according to Bloomberg Intelligence. The rest of the S&P 500 without the Mag Seven is expected to report a 13 per cent increase in earnings and a 5.5 per cent rise in revenue in 2026. 

But there is a divergence within the group’s stock market performance. Nvidia, Alphabet, Meta and Microsoft are considered well positioned for an AI world, and their shares are up between 21 per cent and 33 per cent in 2025. The prospects for Apple, Amazon and Tesla, on the other hand, are less clear, and they are lagging behind badly. 

“It is hard to see the current Mag Seven as the best representation of AI,” Mr Smith said.

So Wall Street has been proposing variations to the theme to capture the real winners. Some have trimmed it to a “Fab Four” of Nvidia, Microsoft, Meta and Amazon. Mr Jonathan Golub, chief equity strategist at Seaport Research, proposes removing Tesla to create the “Big Six”. Others like Mr Ben Reitzes of Melius Research prefer an “Elite Eight” of the Mag Seven plus chipmaker Broadcom, which is now the seventh-largest company in the US by market capitalisation.

But none of those captures the full AI trade. For example, Oracle shares are up more than 75 per cent in 2025 as its AI-related cloud-computing business takes off. And Palantir is by far the top performer in the tech-heavy Nasdaq 100 Index, soaring 135 per cent in 2025 on strong demand for its AI software.

“A company can become too big to ignore,” said Mr Jurrien Timmer, director of global macro at Fidelity Investments. “It could be that as the AI story evolves, new winners take the place of the old winners, even if the previous ones continue to do fine.”

The concept of the Magnificent Seven is nothing new on Wall Street, which likes to create collections of buzzy stocks to simplify the market for investors – from the Nifty Fifty of the 1960s, to the Four Horsemen of the Nasdaq in the dot.com era, to the FAANGs that ruled the days between smartphones and AI earlier this century. But just as those groups were dominant in their moments, they eventually ceded leadership to new names, something that seems destined to happen with AI.

In a sign of how Wall Street is looking beyond the Mag Seven, Cboe Global Markets is launching futures and options based on what it is calling the Cboe Magnificent 10 Index, which includes the original seven along with Broadcom, Palantir and Advanced Micro Devices, Nvidia’s much smaller rival in processing chips.

That index shows how subjective this exercise can be. Cboe’s announcement came on Sept 10, just as Oracle was having its biggest one-day gain since 1992 following a robust forecast that cemented its status as a major AI winner. Oracle’s stock is beating most of the Mag Seven since the start of 2023. But it still didn’t make the cut for the Magnificent 10.

Cboe declined to make someone available to discuss index’s methodology, but said in its press release announcing the Mag 10 that the components were selected “based on liquidity, market value, trading volume and leadership in areas like artificial intelligence and digital transformation”.

Next-gen leaders

Wall Street pros nominated a number of candidates for the next generation of leadership, but some companies were named especially frequently in interviews. Taiwan Semiconductor Manufacturing Co (TSMC) is seen as a critical component of the AI ecosystem along with Oracle and Broadcom. And Palantir is considered one of the few AI software winners at a time when legacy leaders like Salesforce and Adobe struggle against the perception that they are being left behind.

In terms of which stocks are no longer magnificent, Apple and Tesla are mentioned most often. Apple is not generating the same level of growth as the other tech giants, and it is considered well behind in AI. Meanwhile, Tesla’s electric vehicle business has come under heavy pressure as sales dry up and competitors emerge. 

But both still have legions of stock market fans betting they will be there when the time comes. For Apple, the wager is that the iPhone will be the device millions of consumers use to access AI. And Tesla investors hope chief executive Elon Musk’s push for autonomous driving and humanoid robots, which require AI, will lead to future growth.

There is also a lengthening list of industries benefiting from AI, including power generators and other elements of the AI infrastructure build-out, like communications equipment company Arista Networks, memory chipmaker Micron Technology, and storage companies such as Western Digital Corp, Seagate Technology Holdings and SanDisk.

Another challenge in pegging the AI trade is that several key companies are private. OpenAI would likely be on any list of AI winners but is out of reach for most investors, even though it is reportedly in talks to sell stock at a valuation of about US$500 billion (S$645 billion). Anthropic and SpaceX also are not publicly traded.

As AI proliferates, the beneficiaries are likely to change from the companies facilitating the rise, to those offering AI-specific services and products, and finally to businesses using it to improve efficiency and growth. This transition will likely determine AI’s ultimate winners – whatever Wall Street eventually decides to call them.

“As that evolution happens, the leaders of the AI boom may become expensive, their growth and cash flow could stop looking as good, and the trade will start to fray at the edges,” Fidelity’s Mr Timmer said.

“The problem with a concentrated market is that we could have a disruptive change as leadership falls out of favour. We’re not currently at valuation levels that make me scratch my head, but right now we can’t say if the Mag Seven era will end with a benign rotation or a crash.” BLOOMBERG

See more on